TALLIN, J.I.W. Veasureirert cf rental "traits in norr?] & epileptic school chil iren. 8061 'U m iu A 'AI 'asnaFiXg japujg junoiuoqjiT THE MEASUREMENT MENTAL TRAITS IN NORMAL AND EPILEPTIC SCHOOL CHILDREN By J. E. Wallace Wallin. Ph. D., Director Bureau of Special Education, and Professor of Clinical Psychology, Miami University Author The Menial Health of the School Child, Problems of Subnormalily, Experimental Studies of Mental Defectives, etc. MIAMI UNIVERSITY BULLETIN Series XXI, No. £ iA^f^19^®^ Price 75 cents. Send orders to Secretary, Miami University, Oxford, Ohio THE MEASUREMENT OF MENTAL TRAITS . IN NORMAL AND EPILEPTIC SCHOOL CHILDREN By J. E. Wallace Wallin. Ph. D., Director Bureau of Special Education, and Professor of Clinical 'Psychology, Miami University Author The Mental Health of the School Child, Problems of Subnormality, Experimental Studies of Mental Defectives, etc. M T 1 PREFACE In our "Experimental Stitdies of Mental Defectives" we made a detailed analysis of the individual tests in the 1908 Binet-Simon scale based on the test- ing of epileptics while we had charge of the psychological laboratory and clinic of the New Jersey State Village for Epiletpics at Skillman. Since the publica- tion of the above monograph we have published various other studies of the 1908, 1911 and Stanford revisions of the Binet scale. The present study, the experiments of which were carried out during our connection with the Skillman institution, gives the results of a comparative investigation of epileptic and normal children by means of about a dozen inde- pendent psychological group tests. The publication of this monograph has been greatly delayed, partly because of our heavy schedule of work, leaving no time for working up researches except during off-hours, and partly because of the great difficulty during the last few years of securing publication of original experimental contributions, owing to the high cost of printing. The manu- script has been ready for some years. The value of the experimental results, however, is not impaired by the delay of publication, particularly in view of the fact that since this study was made not a single extensive experimental investi- gation has appeared on the psychology of epileptic and normal children. The experiment here described is based upon the use of one of the first extensive sets of group tests devised for the measurement of intelligence, many of the tests being first used by the writer in an experiment under the auspices of the Oral Hygiene Committee of the National Dental Association, on a squad of children receiving special dental treatment. The use of group tests of intelli- gence did not become popular until the Army Alpha and Beta appeared. Since then scores of so-called group tests of intelligence have appeared in quick suc- cession, embodying numerous refinements and improvements in the character of the tests used, and in the administrative and scoring technique. Literally millions of children and adults are now being tested and rated every year by means of these tests. The tests used in this experiment were administered to 106 subjects in five successive sittings. All in all, we examined 5,839 record blanks; 4,179 of these were handed in by the normal children, and 1,660 by the epileptics. This includes the deferred recall of a number of tests (A2, Fl, F3, Gl, G3, G4 and K4) after the lapse of one month. Exclusive of the latter tests, 3,758 papers were handed in by the normals and 1,494 by the epileptics, or a total of 5,252. The number of record sheets in each test from A to J (exclusive of H) was somewhat over 360 for the normals and about 145 for the epileptics. It would be impossible within the limits of the time and space here available, to analyze this mass of data in minute detail. We have therefore taken pains to publish the original tables (which, however, contain only the quantitative tabulation), in unabridged form, so that they may be accessible to workers who may want to verify our results or who may want to pursue the analysis further than we have been able to do. However, while the complete data are given for each one of the epileptics studied, only averages are given for the normal pupils. We have separately tabulated the results for groups of "bright," "average" and "dull" public school children arranged according to school grade. These data are not included in the present study. They will be made available to anyone who has the time to work them up. J. E. W. W. 3 CO XT ENTS Page Preface 3 I. Introduction 5 II. Memorizing Visually Presented Numbers 22 III. Rapidity of Thinking 42 IV. Addition Test 53 V. Controlled Association 66 VI. Attention and Perception 79 VII. Observation 88 VIII. Memory of Logical and Illogical Associations 112 IX. Visual Imagination 125 X. Word Construction 132 XI. Sentence Construction 140 XII. Recognition 146 XIII. Speed of Motor Reaction 157 XIV. General Comparisons and Conclusions 162 4 I. INTRODUCTION The following research was carried out upon epileptic and normal school c hildren during five successive months. The normal children, so-called, were pupils enrolled from the second grade to and including the eleventh grade of the public school in Hopewell, N. J. (this school did not contain a twelfth grade, or a fourth high school year), while the epileptics were attending the school in the New Jersey State Village for Epileptics at Skillman. The principal and teachers in the school at Hopewell were requested to select from each grade above the first for the investigation two dull boys and two dull girls, two average or medium boys and two average girls, and two bright boys and two bright girls, making a total of 120 children. The teachers were asked to rate the pupils not only on the basis of the proficiency they had shown in the school work, but also on the basis of the teacher's independent estimate of their intelligence. Actually 99 children were selected by the school staff as meeting the specified requirements. Twenty-three of these children, however, had to.be rejected from the final tabulation, either because they failed to attend regularly on the test days or because they moved from the district. Xo one among the normal pupils was included in the final sum- mary who was absent on more than one test day (although two papers were lacking for three subjects in three different tests, and two papers for three subjects in test K4). Unfortunately quite a number of the normals missed one of the sittings, the number of misses varying from 4 to 19 in tests A to I. Some of the pupils came late and thus missed some of the tests, while a few evidently failed to pass in some of their papers. (The papers were collected at the conclusion of each test.) Of the 76 normal pupils whose records could be included in our tables, all of whom were white except one colored girl, 39 were girls and 37 boys, approximately an equal number of boys and girls. The pupils were classified as average, dull or bright by the school staff before the beginning of the experiment in January and again after the con- clusion of the experiment in May. The classifications were the same in Jan- uary as in May for all the pupils except three. The classification of one of these was changed from average to bright, one from bright and one from dull to average. According to this classification, 28 of the pupils were "bright" (12 boys and 16 girls), 29 were "average" (16 boys and 13 girls), while 19 were "dull" (9 boys and 10 girls). The distribution of the bright, average and dull pupils in each age is given in Table 1. 5 6 The Measurement of Mental Traits in- table i Number of Dull, Average and Bright Pupils among the Normal Children in each Chronological Age. Dull Average Bright Age Boys Girls Both Boys Girls Both Boys Girls Both 7 1 o 1 0 1 0 2 2 8 1 1 2 5 1 6 2 1 3 9 1 1 2 1 1 2 1 2 3 0 1 1 2 1 3 2 1 3 2 2 4 1 3 4 0 3 3 12 1 2 3 2 0 2 2 1 3 13 0 0 0 2 1 3 2 1 3 2 2 4 0 3 3 2 2 4 15 0 0 0 1 0 1 1 2 3 16 1 0 1 0 2 2 0 1 1 17 0 1 1 1 2 0 0 0 Total 9 10 19 16 13 29 • 12 16 28 According to the arbitrary age-grade standards (Grade I, 6 years; Grade II. 7 years; Grade III. 8 years, etc.), 21 of the pupils (6 girls and 15 boys) were accelerated in amounts ranging from one to two years, while 35 pupils were retarded (23 girls and 12 boys) from one to four years. But the aver- age amount of pedagogical retardation was only .39 year ( .08 for the boys and .68 for the girls) . The average chronological age of the boys was 11.47 (varying from 7 to 17.08 years), while their average school grade was 6.2 (median 6) ; and the average chronological age of the girls was 12.3 (varying from 7.25 to 17 years), while their average grade was 5.9 (median 6). The corresponding figures for the entire group are 11.80 and 6. It appears from the pedagogical status and from the teachers' estimates that we are dealing with a good group of normal children, perhaps somewhat superior to the average. Probably not more than one pupil in the entire group approximated the feeble-minded status.1 The experiment began with the 39 ablest epileptics in the school depart- ment of the State Village. This number was subsequently reduced to 30, 16 boys and 14 girls (all white), owing to the elimination of 6 pupils who were unable to perform the tests because of their low mental status, owing to the demise of one boy and owing to the rejection of the records of two pupils who were very irregular in attendance because of convulsions. Although all of the epileptics whose records are utilized were present on all of the test days, from 3 to 6 missed one or more of the tests in a given sitting, owing to the fact that they had just had a convulsion before the testing began, or had a convulsion during the course of one of the tests, while an average of one epileptic was unable to take one or more tests in two sittings for the same reason. The average chronological age of the epileptic boys was 14.56 'For a representative normal group selected on a somewhat similar basis see J. E. Wallace Wallin, Psycho-Motor Norms for Practical Diagnosis, 1916, p.9ff. Normal and Epileptic School Children 7 (ranging from 10.58 to 17.83); of the epileptic girls 15.76 (from 10.83 to 24.25) ; and of the entire group 15.11. (See Tables 2 and 3.) The epileptic boys averaged 3.09 years older than the normal boys; the epileptic girls 3.45 years older than the normal girls, and the epileptic group 3.31 years older than the normal group. Only 26.6% of the epileptics as against 63.1% of the normal children were 12 years or less in age. On the other hand, 73.3% of the epileptic but only 36.8% of the .normal children were over 12 years old. Whatever advantages attach to greater age or physiological maturity were decidedly in favor of the epileptic group. Although the epileptic children thus averaged considerably older chronologically than the normal children, they were inferior or less mature mentally. This is shown both by their pedagogical status and by their mental age. The average amount of pedagogical retardation for the epileptic boys was 7.0 years, for the girls 7.3. and for the epileptic group 7.2, based on the standards given on page 6. (Xo epileptic, however, was considered as over 17 years in determining the amount of pedagogical retardation.) This is many times larger than the corresponding figures for the normal children, p. 6. All of the -epileptics were retarded in amounts ranging from a minimum of two years for the girls (one case) and 4 years for the boys (two cases), to 10 years plus for the girls (2 instances) to 9 years for the boys (4 instances). The decidedly more favorable figures on acceleration and retardation for the normal children are given on p. 6. Thirty per cent of the epileptic boys were doing first grade work; 30% second grade, and 40% third grade work; 28.5% of the epileptic girls were in the first grade, 35.7% in the second and 35.7% in the third; 30% of the entire group were in the first grade, 30% in the second, and 40% in tne third grade. Xo one did work beyond the third grade. On the other hand, only 18.4% of the public school children were in grades II and III (none in Grade I), while 81.5' , were in Grades IV to XI (Third High) ; 56.5% were in Grades II to VI, and 43.4% in Grades VII to XI. The average school grade for the epileptic girls was 2.3 (median 2.), for the boys 2.1 (median 2.) and for the whole epileptic group 2.1 (median 2.). The normal group aver- aged about four grades higher. The average Binet-Simon age (1908, the writer's guide1) of the epileptic boys was 10.1 (median 10.2), of the epileptic girls 9.7 (median 10.4), and of the epileptic group 9.9. (median 10.4). On the supposition that the aver- age mental or intelligence age of the normal group corresponded approximately to their chronological age, the normal boys averaged 1.37 years higher mentally than the epileptic boys, the normal girls 2.61 years higher than the epileptic girls, and the normal children 1.9 years higher than the epileptic 'The Binet tests were given in 1910 and 1911 before the 1911 revision had appeared, hence it is impossible to compute the results in terms of both the 1908 and 1911 scales, as we have done elsewhere for other subjects. 8 The Measurement of Mental Traits in children. Relatively the epileptics are decidedly more deficient pedagogically than mentally. Twelve and five-tenths per cent of the boys (2 cases) graded B.-S. VIII years,1 18.7% (3 cases) IX years, 50% (8 cases) X years, and 18.7% (3 cases) XI years; 28.5% (4 cases) of the girls graded VIII years, 14.3% (2 cases) IX years, and 57.1% X years. In other words, 20% (6 cases) of all the epileptics reached a mentality of VIII, 16.6% (5) of IX, 53.3% (16) of X and 10% (3) of XI. The average amount of Binet-Simon retardation was 4.3 years for the boys, 6.0 for the girls and 4.8 for both sexes, the retardation being determined in every case by subtracting the exact Binet-Simon from the exact chrono- logical age (both expressed in decimals), a method known to be faulty for older adolescents. One boy and one girl were retarded from .3 to .5 year; 1 boy and 2 girls from 1.0 to 1.4 years; 2 boys and 1 girl from 2. to 2.6 years: 4 boys from 3.0 to 3.8 years; 2 boys and 1 girl from 4.0 to 4.3 years; 2 boys and 3 girls from 5.0 to 5.9 years; 1 boy and 2 girls from 6.0 to 6.6 years; 2 boys and 1 girl from 7.0 to 7.4 years; 1 boy and 2 girls from 8.1 to 8.5 years; and one girl 10.5 years. Five epileptics had a B.-S. retardation of less than one and a half years, eight had a retardation less than two and a half years, and 12 less than four years. Twenty-two had a retardation of three and over; 18 of four years and over; 15 of five years and over; 10 of six years and over, and 7 of seven years and over. What percentage of the epileptics was feeble-minded? On the basis of the hypothetical three-year retardation standard, 75% (12) of the boys were feeble-minded, 71.4% (10) of the girls, and 73.3% (22) of the whole gfoup. On the basis of the four-year retardation standard, 60% of the entire group would be feeble-minded. On the basis- of the IX-year upper threshold of feeble-mindedness,2 only 31.3$ (5) of the boys, 42.8% (6) of the girls and 36.6% (11) of the whole group were feeble-minded. None of these arbitrary standards, or others, however, can be rigidly applied to the diagnosis of feeble-mindedness in epileptics. For example, many epileptic children retarded only one year mentally will, no doubt, have to be diagnosed as feeble-minded, because it is very probable that they will fail to develop mentally, or they will slowly deteriorate. The tendencies toward arrest or deterioration for this group of children can be seen from an inspection of Tables 2 and 3. These tables are based on data sent me by Dr. D. F. Weeks, Superintendent of the New Jersey State Village for Epileptics, after four and a half years of further training. Twenty-two of the children were in resi- dence at the institution at that time. The average pedagogical grade in the academic work of the nine who were still in school was 2.7 as against 2.1 at the time of the first report, an average pedagogical gain of only half a year ■Age VIII includes everything from 8. to 8. 8, and similarly for the other age designations. 2See J. E. Wallace Wallin, Problems of Subnormality. 1921, Chapter II. Normal and Epileptic School Children 9 in four and a half years. Sixteen, or 72.7%, of the 22 were reported as having retrograded pedagogically, only two were said to have improved pedagogically (although five of the nine in the school were classified in a higher grade), while four were recorded as stationary. Mentally 17 were reported as having retrogressed, two as having improved and three as having remained stationary. At the time of the examination the writer estimated that 64.3% (9) of the girls, 62.5% (10) of the boys, or 63.3% (19) of the whole group were feeble-minded, all belonging to the higher grade of feeble-mindedness. On the other hand, according to the estimates sent by the institution over four years later, all of the patients in residence were feeble-minded, assuming that the word "moron" was used as synonymous with the highest grade of the feeble-minded. 10 The Measurement of Mental Traits in- table 2 Data on Epileptic School Children at New Jersey State Village at Skillman. During Experimental Period. G. G. G. B. G. G. G. B. B. B. B. B. B. B. G. 12.33 17.16 12.08 15.41 17.50 16.50 13.58 13.00 10.83 15.75 16.00 12.41 10.58 16.50 10.91 11.75 18.58 21.16 14.00 13.00 16.08 16.41 16.91 16.75 15.16 17.83 12.41 14.00 15.00 24.00 8 2 8.6 8.8 8.8 9.0 9.2 9.2 9.3 9.4 8.3 10.1 10.2 10.2 10.4 10. 4 10.4 10.4 10.6 10 6 10 6 1st 2nd 1st 2nd 3rd 3rd 1st 1st 3rd 2nd 2nd 1st 1st 2nd 1st 1st 2nd 3rd 3rd 1st 3rd 3rd 3rd 2nd 2nd 3rd 3rd 3rd 3rd 2nd Type of Convulsions during period of Experiment Grand and petit mal Grand and petit mal Grand and petit mal Grand and petit mal Grand and petit mal Grand and petit mal Grand mal only Grand and petit mal Grand and petit mal Grand and petit mal Not in the village Grand and petit mal None Not in the village Grand and petit mal, nocturnal Grand mal only Not in the village Grand and petit mal Not in the village No convul- sions during period Grand and petit mal Grand and petit mal Grand and petit mal Grand and petit mal Not in the village Grand and petit mal Not in the village Not in the village Not in the village Grand and pelit mal Approximate Frequency per month of Convulsions at time of Experiment 61.8 6 9 19 35.4 9 15.6 15 65 9.7 9.8 None 20 5.4 51 Unknown •37 11 18.1 19.3 Onset of Convulsions Age of incidence 10 months 9 years 11 months Wi years 13 years 1 year 8 month 3J4 years 6 years 5 years 5 years 8 years 8 months 10 months 13 years 2V2 years 7 years 2 years 6 months 8 years 5 years 17 months Etiological or pajho- logical factors, so far as known Dentition Unknown Indigestion Dentition Unknown Unknown Unknown Unknown Indigestion Unknown Pertussis Unknown Dentition Dentition Puberty Unknown Trauma of head . Dentition Unknown Unknown Overdose of ether Dentition Extent of use of bromides during ex- perimental period. To control series To control series To control series To control series To control series To control series To control series To control series To control series To control series To control series Unused To control series To control series To control series Unused To control series To control series To control series To control series To control series To control series Normal and Epileptic School Children 11 TABLE 3 Data on Epileptic School Children at New Jersey State Village at Skillman. History Four and Vi Years Subsequent to Experimental Period. Subject No. 1 Type of Convulsions Frequency of Convulsions per month Present Schoo Grade i Improved or retrograded pedagogical! y? Improved or retrograded mentally? Present classifications ( Imbecile, moron, borderline, backward, etc.) 108 Grand and petit mal 12 2nd Improved Improved Imbecile 95; Grand and petit mal 2.4 2nd Retrograded Retrograded Middle imbecile 79 Grand and petit mal 10.6 4th Retrograded Retrograded Imbecile 103] Grand and petit ma! 11 . 1 Not in school Retrograded Retrograded Low imbecile 87 Grand and petit mal 32 3 Not in school Retrograded Retrograded Imbecile 100 Grand and petit mal 6.1 3rd Retrograded Retrograded Middle grade imbecile 86 Grand mal only 10 Not in school Retrograded Retrograded Imbecile 93 Grand and petit mal 19 Not in school Retrograded Retrograded High grade imbecile 102 1 Grand and petit mal 9.6 2nd Retrograded Retrograded Middle imbecile 83| Grand and petit mal 9 2 Not in school Stationary Stationary Moron 78 Not in Village 91 Grand and petit mal 26 Not in school Retrograded Retrograded Middle imbecile 81 1 Grand and petit mal 33.2 Not in school Stationary Retrograded Moron 106 Not in Village 104 Gd. and pt. malnoctl. 16 2nd Retrograded Retrograded Moron 99 Grand mal 43.5 Not in school Retrograded Retrograded Imbecile 105 Not in Village lo- Grand and petit mal 31 Not in school Retrograded Retrograded Low imbecile ss Not in Village 97 No Convulsions Not in school Stationary Stationary Moron 96 Grand and petit mal 24.9 2nd Retrograded Retrograded Middle imbecile 98 Grand and petit mal 13.3 Not in school Retrograded Retrograded High grade imbecile 85 Grand and petit mal 16.7 5th Improved Improved Moron 98 Grand and petit mal j 7.6 3rd Stationary Stationary High grade imbecile 80 Not in Village 90 Grand and petit mal 23.9 Not in school Retrograded Retrograded Imbecile 84 Not in Village 77 Not in Village 82 Not in Village Retrograded 1 101 Grand and petit mal \ 6 Not in school Retrograded Middle imbecile At the time of the experiment eighteen of the epileptics suffered from both the grand and the petit mal type of convulsions, two from the grand mal type alone, two did not have any seizures, while the data are not recorded for eight of the subjects. The later record indicates that 19 were suffering from grand and petit mal seizures, two from grand mal (the same subjects as before), one had no seizures (this girl had no seizures in 1911), while the data are lacking for eight subjects. The seizures were of the mixed type in the vast majority of cases for which the data are supplied. The details in regard to the frequency of the convulsions are accessible in the tables. Our group of epileptics shows the usual variation in the frequency of the attacks. Bromides to abort seizures were not used at the institution except in the case of serial attacks (status epilepticus). The subjects accordingly were comparatively free during the experimental period from bromidic hebetude. The convul- sions began at or before the age of 8 years in 19 cases, after the age of 8 in 3 cases, while no data are given for eight subjects. The onset occurred before the age of 6 in 15 cases, and before the age of 3 years in 11 cases. 12 The Measurement of Mental Traits in In addition to the above facts, we recorded data on the condition of each epileptic during the different sittings. The following are some of the more significant items:1 No. 108 was in a worse mood than usual during most of the second sit- ting. She did not do so well in some of the tests during this sitting. No. 95 had convulsions the night before the second sitting, and only made feeble efforts in the tests. She was out of the room part of the time. She suffers from the after-effects of convulsions. No. 79 was below normal during the first sitting, having had a convul- sion earlier in the morning. During the fifth sitting he had a convulsion just before test C was to be given, but partook in some of the later tests. Xo. 103 had a seizure 10 minutes before the tests started in the second sitting, and did nothing in most of the tests on this day. No. 86 said he felt good during the second sitting. On the day of the fifth sitting he had had two grand mal seizures. He did poorer during this sitting in practically all the tests in which he had previously made scores. No. 93 was reported sub-efficient during the first sitting, owing to the fact that he had had two convulsions earlier in the morning. His record in the tests was perhaps somewhat below normal during this sitting. No. 102 had a convulsion during the latter part of the second sitting. She appeared drowsy a short time before and did nothing after the convulsion. During the third sitting she said she was sick. She had a convulsion before coming to the class and was reported to have been stupid all week. She did poorer in most of the tests during these sittings. No. 105 was somewhat stupid and refractory during the third sitting and more or less stupid during the fourth sitting, but did as well as usual in some of the tests. No. 96 had a convulsion before the tests began during the first sitting, and had a convulsion after test A had been given during the fourth sitting. She returned after test D had been given. The aim of the psychological experiment was to measure under controlled conditions the comparative amount of change or development which might occur in various mental traits among various grades (bright, average, dull) of normal pupils of different ages and school grades and among epileptic school children, during the same intervals of time and under the same experi- mental conditions. To be able to measure mental changes we must have at hand a set of mental tests each of which is so simple of comprehension and execution that it can be given to children of various ages, each of which must be so scored that the youngest children will be able to make a score above 'Data on the condition of the normal children during each sitting were also gathered. We can- not take the space to detail these. Normal and Epileptic School Children 13 zero, while the oldest or the ablest pupils must not be able to make a perfect core, and each of which must be constructed in a successive series of tests of he same degree of difficulty.1 It is evident that if each successive test increases in difficulty, the mental development will be minimized or neutralized by the increasing difficulty of the successive tests themselves, while, on the other hand, if the successive tests decrease in difficulty the mental improvement will be exaggerated. In an ideal set of serial tests designed accurately to measure mental growth or improvement the successive tests of the same series must be absolutely uni- form in difficulty, but it is difficult to construct such uniform series, and this is true in some tests more than in others. The difficulty is increased by the fact that, owing to the influence of the factors of familiarity, practice and memory, it is not permissible — or at least it is inadvisable — to use precisely the same test material in each of the successive tests. However, by intro- ducing certain changes in the arrangement of the same material, or by using new material of the same degree of difficulty in each of the successive tests, we shall be able to combat, in part at least, the influence of the factors of familiarity, practice and remembrance which inevitably make themselves felt in repeated reactions to identical test materials. The introduction of altera- tions, however, may make it difficult to secure absolute uniformity in the different successive tests, particularly when entirely new materials are intro- duced, as we shall see later. But the inequalities thus produced are less objectionable than the repeated employment of exactly the same test ma- terials. In fact, it would be impossible to employ the same materials in some of our tests (e. g., the spontaneous association and antonym tests) during the five sittings in which the tests were conducted.2 In the following tests the same materials were used, but in a rearranged form, in all of the successive sittings: immediate reproduction of memorized three-place digits (A), and speed and accuracy of perceptual discrimination (E). The same letters were also used in some of the word building tests (I). The arrangement of the digits in the later addition tests (C) was secured by a systematic rearrange- ment of the digits in the earlier tests. In the following tests new materials were introduced in each successive test: spontaneous association with sup- plied key-words (B), the giving of antonyms to match supplied key-words (D), range of visual apprehension (F), immediate recall of sequents (or antecedents) in presented paired associates (G), visual imagination with ink blots (H), sentence construction with three supplied words (I), and recogni- tion of illustrated post cards (K). The detailed nature of the test materials used may be seen by examining the following chapters in which the materials are described, or those used in the first sitting reprinted. ■We have elsewhere discussed these and other conditions of a series of tests of this nature: Dental Cosmos, 1912. p. 545f. 2The tests were constructed in six series, but it was impossible to conduct the last sitting. 14 The Measurement of Mental Traits in Practically all of the tests used came within the lower limit of ability ot the least capable or the youngest normal pupils who were tested. Only in one test. J, did as many as seven subjects fail in all five sittings, while 16 failed in the first sitting. Even in this test, however, some of the youngest subjects scored in some of the sittings. In only one other test, C, based on the correct additions only, did more than one subject fail in the five sittings. Three failed in all the sittings in this test. In six tests there was one failure in all five sittings, but this was made by the same subject, a seven-year-old boy in the second grade, classified as dull, who was more or less restless and who usually applied himself very poorly. A number of subjects, particularly the younger ones, grew more or less inattentive during the course of the test- ing, i. e., during the last two or three sittings. A few made little effort in some of the tests in the later sittings, while a few attempted to interfere more or less with other pupils. These factors resulted, no doubt, in slightly reduc- ing the general averages. There was not a single failure in any sitting in three tests (A, E and L>, while in five others less than three pupils failed in any one sitting. On the other hand, the tests were administered under such rigid time limitations that only a few of the ablest pupils were able to finish the test or make a perfect score. Only six different subjects made perfect scores in any one sitting in any of the tests (all in the three last sittings), and no subject made perfect scores in more than one sitting of the same test. Only one subject — a 16-year-old girl classed as "bright" — made a perfect score in one sitting each of two tests, namely in the antonym (D) and recognition (K) tests. Perfect scores were made in only three tests, the antonym test (D), by one subject, the paired associates tests (G), by four, and the recognition test (K), by two. All of the subjects who made these perfect scores were 14 years old or over, except one, who was 11 years. All were in the seventh grade or above except the 11 -year-old (sixth grade). Three of these pupils were classed by the teachers as bright, one as average and two as dull. One of the latter was the 11-year-old boy who may have been rated dull relatively to older pupils in the sixth grade. The responses to all the tests required the use of a lead pencil, since they were designed as group tests and all were so administered. Had the tests been given individually or clinically, so that the responses could have been made verbally instead of in writing, in the tests which permitted of verbal responses, the scores would, of course, have been higher, at least in some of the tests and especially so far as concerns the younger children and many of the epileptics, whose responses were impeded by the inability to write rapidly and by uncertainty as to the spelling of words. To minimize this handicap the children were urged to write as rapidly as possible irrespective of the beauty of the penmanship and irrespective of the correctness of the ortho- Normal and Epileptic School Children 15 graphy. The pupils were told that misspellings did not count, and that, therefore, they should not lose any time pondering over the spelling of uncer- tain words, but spell as well as they could. The tests were given at intervals of twenty-eight days between January 17th and May 9th. We should have preferred giving them at longer inter- vals, every third or sixth month, but the time available for the experiment made this impossible. We would emphasis, however, that the tests were given to our two groups at precisely the same intervals of time — the normal pupils on Tuesdays and the epileptics the following day. Moreover, both groups were tested during the same part of the day, the normal pupils in three different groups between 9 A. M. and 12 M., and the epileptics in two groups between 9 and 11 A. M. The weather conditions were good on all of these mornings, except that the sky was cloudy during the February test, and the weather was rather warm and sultry during the May test. The epileptics seemed to be in a better condition during the May sitting than during any other sitting. . The time required to give all the tests to one group of children was ap- proximately one hour. The same rooms, well lighted and free from noise — both recitation rooms, the one in the public school and the other in the school for the epileptics — were used in all the sittings, and the pupils occupied their assigned seats on all of the test days. The tests were all administered by the writer, under rigidly uniform conditions. The tests were timed by means of a stop watch. The instructions were given in a clear, loud tone of voice, easily audible in all parts of the room, while the materials displayed in the memory (A) and observation (F) tests were easily visible from any seat in the room occupied by the pupils. Two assistants aided in the distribution and collection of the papers and served as ' watchers," in order to lessen the temptation of the pupils to copy each other's papers. The pupils were explicitly warned not to copy and no copy- ing was observed. The general procedure which was followed in giving the tests is indicated by the following directions (the statement is slightly modified to meet the present needs) which accompany the test materials.1 GENERAL DIRECTIONS Explain to the children at the outset that you are going to give them a series of tests or games, and that you want to see just how well they can do in the tests. Urge them before giving each test to do their very best; to work, to write, to think just as well and just as fast as possible. But do not 'The materials, including the six complete series in the different tests, can be obtained from C H. Stoelting Co., 3037 Carroll Ave., Chicago, 111., with the exception of Fl to 6, K4 and L5. Five of these tests ( A. B, C, D and E I were originally prepared for an experiment given under the auspices of the Oral Hygiene Committee of the National Dental Association: J. E. Wallace Wallin. Experimental Oral Euthenics. Dental Cosmos, 1912 , 54:404-413; 543-566. 16 The Measurement of Mental Traits in excite them. Have them work naturally, without undue constraint, fear or haste. Avoid suggesting the object of the entire investigation as such. Let the work appear as a phase of the regular routine of the day, or better still, as a series of interesting contests or games. But the subjects should do their best. The experimenter must be tactful. Explain that you are going to place blanks (sheets of paper) on the desks upside-down ; that when you say "ready" they must take their pencils in one hand and at the same time take hold of the lower left corner of the papers with the other hand; that the very moment you say "turn" or "turn your papers" they must rapidly turn the papers over sidewise (always sidewise), and begin to write at once; and that when the time is up you will say "stop." They must then cease writing at once ; they must not write another letter, but instantly turn the papers over again. Make sure that all the subjects are ready before the starting signal ("turn") is given, and that everybody turns the papers the moment the "stop" signal is given. Explain that no pupil should stop to ask questions or to sharpen pencils while the tests are being given. Each pupil should have two pencils, well sharpened. After the pupils have turned up the reverse side at the close of the first test, have them write at the top of the page: full name. sex. age (verify the age from the official records), date and hour of test, condition during tests (well, fresh, tired, sick, pains, etc.), school, and grade. At least the name must be written on the subsequent test blanks. Absolutely uniform conditions must be maintained throughout all the tests; same seating, same instructions, same experimenter, if possible, same hour of day, uniform absence of distracting stimuli. The time set for each test must be minutely observed. The experimenter will always start the stop watch the very moment the cardboards are displayed (Al to A6, Fl to F6, K4), or at the moment the starting signal is given ("turn"). He will stop the watch the moment the "stop" signal is given. The pupils subjected to the tests should be especially urged to be present on the days on which the tests are scheduled. In case any are found to be absent, they may be tested during the corresponding hour on the first day of their return, but note must be made of this fact. (None was tested except during the regular sittings in this investigation.) Pupils should be so seated as to avert possibilities of copying. They should be told that if they do copy the papers will be discarded. One or two helpers should assist the examiner to distribute the papers and keep tab on the pupils so that no copying can be done. ORDER OF GIVING TESTS One from each series should be given at each sitting. The order of the successive sittings is indicated by the numbers 1, 2, 3, 4, 5, and 6, printed at the bottom of the test blanks after the letters A, B, C, etc., 1 coming during Normal and Epileptic School Children 17 the first sitting. 2 during the second sitting, etc. The letters A, B, C, D, E. F. G, H. I. J, K. and L indicate the order in which the different kinds of tests are given in the same sitting. A coming first, then B. then C, etc. The tests follow each other as fast as they can be given. The following scheme will indicate the order: During the first sitting: give Al, Bl. CI, Dl, El, Fl, Gl, HI, II, and Jl. During the second sitting: A2, B2, C2, D2, E2, F2, G2, H2, 12, and J2. The sittings may come at intervals of a month, two months, a half-year, a year. etc. Some may be given before the introduction of a special factor (e. g.. dental treatment, administration of thyroid extract) and others at various subsequent intervals. The directions for administering the individual tests are given in con- nection with the discussion of each test in the subsequent chapters. EXPLANATION OF TABLES The following explanations apply to all the tables which contain the original data, although some of the functions have been worked out for the first seven tests only. No., number of subjects included in the averages. Since some of the subjects missed one of the sittings (and a few missed two) the number of subjects given in the first column of the tables for the normals does not hold for all of the sittings. Su., subject; Ae., age; Sx., sex; B., boys; G., girls. ■},,•;,, l» it and £, percentages of efficiency shown in the first, second, third, fourth and fifth sittings, respectively, given on the following dates for the public school children (and on the next following day for the epileptics): 1 2 3 4 5 % % % % c/o Jan. 17. Feb. 14, March 14. April 11, and May 9. We shall refer to these successive test days as sittings, and to the per cents made during any sitting as scores or percentage scores. The figures in the tables for the normals are averages except for age 16, boys, where there is only one subject. The averages for all the functions in all the tables were obtained by summating all the individual scores, whether for a given age or for all the ages or for either or both sexes, and dividing by the total number of cases. The units for scoring the different tests differ widely according to the nature of the material used in each test. We have converted the scores in nearly all the tests into percentages, by giving a percentage value for each unit or part unit of work accomplished in the different tests, based on 100% as a perfect record. While there are certain advantages in reducing the re- sults to a percentage basis, it should not be inferred that a given percentage M T 2 13 The Measurement of Mental Traits in signifies the same degree of efficiency in the different tests. An inspection of the results in the tables of the different tests for the children of the same age shows that this is not so. A 50r^ efficiency in one test may be equal to a 75% efficiency in another test. Let us repeat, then, that there is no value equivalence between the same percentage scores in the different tests. It is evident that any one who may use the tests need not score the results in terms of per cents, as we have done, which involves a considerable amount of labor. We have found the per cent method convenient for scoring partial credits or making deductions for omissions or partial errors, and for expressing the re- sults in a combined qualitative-quantitative score. I%2, average of the per cents for the first and second sittings (in test B, 2 and 3 are averaged), , same for sittings 4 and 5; ]'/', same for sittings 1 to 5. In averaging the scores for different months whenever the subject was absent during one of the sittings the score in the next preceding or next fol- lowing sitting was substituted (except in test B. in which the January results could not be used, because of difference in time). Thus January and March were averaged when February was lacking, or March and April when May was lacking, or March and May when April was lacking. The irregularities produced by this procedure are perhaps usually less objectionable than using the results from a single month for an average. Owing to occasional absences the results for J%5 are sometimes based on less than five sittings. In all cases the figures in the columns V ' are based on the total score for all the monthly tests divided by the actual num- ber of sittings in which a given subject was present. The age averages for each sex or the two sexes in V "' were not obtained by dividing the totals of all the scores by the number of subjects inc'.uded in the given age, but by the number of separate scores included in the total. The final averages were obtained by dividing the grand total of all the monthly scores by the total number of individual scores made by all the subjects. There are two methods by which the improvement made in each test from sitting to sitting can be determined. First, we may calculate the differ- ence between the average scores made in each test. Thus, the difference be- tween the averages of the first and the second sittings for all the normals in Table 4, is 5.4% (a gain) ; between the second and third sittings .4% (gain) ; between the third and fourth sittings 4.4r( (gain); between the fourth and fifth, -4.5% (a loss); between the average of the first two sittings and the average of the last two 5.2% (gain) ; and between the first and the fifth sit- tings 5.7% (gain). This supplies a simple expression of the improvement in terms of absolute units, but it is not an entirely satisfactory method be- cause the relative amount of improvement made by different subjects or different groups of subjects depends less upon the absolute units of gain made than the amount of gain made in proportion to the degree of efficiency — the Normal and Epileptic School Children 19 size of the scores — shown in the initial test. To express the relative amount of gain we have employed the following indices of improvement: f\\ this represents the index of improvement shown in the second sit- ting as compared with the first. It is obtained by dividing the percentage score made in the first sitting into the percentage score made in the second sitting. To obtain the percentage of improvement made in the second sit- ting substract unity (1.0) from the quotient. To obtain the amount of loss substract the quotient from unity. Thus the improvement made by subject 84 in the second sitting of the memory test compared with the first sitting, Table 5. is 85f/c, obtained thus: Sitting 1 (23.3) divided into sitting 2 (43.37 )=1.85. 1.85—1.00 (unity)=.85 (=per cent of gain). On the other hand, subject 104 lost 507c: Sitting 1 (13.27) divided into sitting 2 (6.6%) =.90! 100 (unity)— .50=. 50 (=per cent of loss). An index of 1.0 (unity) indicates that the subject neither gained nor lost. Such a record is shown by subject 82, in sittings 1 and 2, Table 5: 36.6%-=-36.67c=100. 1.00—1.00=0 (=neither loss nor gain). Accordingly in the relative value of the score in 2 is determined by the size of the score in 1, regarded as unity, whence the index indicates the relative amount by which sitting 2 is larger or smaller than sitting 1. The following indices of improvement, j f, / and / r\ are obtained similarly, by dividing the score in the second sitting into the score in the third, the third into the fourth, and the fourth into the fifth. Owing to the fact that a considerable number of subjects missed one sitting (and a few two sittings) it has frequently been difficult to compute the index. Rather than leave the index blank in such cases, it seemed de- sirable to substitute the score in the next sitting, either the one preceding or the one following. Thus, the January score would be divided into the March score when the February paper was missing, the February score would be divided into the April score when the March paper was missing, etc. It must be emphasized that this introduces certain unavoidable errors. In /£§, tne average of the scores in sittings 1 and 2 is divided into the' average of the scores in sittings 4 and 5 (the average of these sittings being given in columns 13 and 14). As we have already remarked, occasional absences made it necessary to substitute the score from an adjacent month to derive an average. Accordingly sittings 1 and 3, and 3 and 5 have some- times been averaged. * • The index / f, in the sixteenth column, shows the gain or loss made in the last sitting compared with the first. Infrequently it was necessary to divide 4 into 1, owing to absences in the fifth sitting. The division in all the indices was carried to two decimal places. It should be pointed out that the indices in the tables for the public school cases have not been computed from the percentage scores given in these tables, 20 The Measurement of Mental Traits in because these scores represent averages. The indices in these tables were ob- tained by finding the index for each subject separately, adding the indices thus obtained for all the subjects of a given age and then dividing the sum by the number of subjects who were present during a given sitting. The index so obtained b not identical, of course, with the index obtained by using the average scores as dividends and divisors. Thus the index / j- for the two 7-year-old boys in the memory test, Table 4, is 1.10. This figure was ob- tained by adding .70 and 1.51, the separate indices for the two subjects, and dividing by 2. The index obtained, however, by dividing the average score for these two subjects in sitting 1 (31.6) into the average score in sitting 2 (25.0) is .79. One reason why the average scores of a given age cannot be employed as a basis for computing the index is the fact that all the subjects were not present during all the sittings, whence the number of subjects aver- aged in one sitting would differ from the number averaged in the next sitting. It would appear that the two methods of expressing the improvement, namely, by the absolute difference between the percentage scores and by the index, should always harmonize. A detailed examination of the tables shows that frequently they do not, either in the direction or in the amount of the change. Thus to take an extreme case: there was a loss for the boys in the second sitting of the addition test for the epileptics. Table 16. amounting to 21%, according to the second index of improvement (/ .79. at the foot of the column). On the other hand, according to the difference between the percentage scores in sittings 1 and 2 there was an actual gain in efficiency in the second test, amounting to .8% (3.0% vs. 3.8%). What is the explana- tion of this discrepancy? It is due to the fact that it is impossible to cal- culate any index of imp:ovement when the scores are zero in both of the sittings, or when the scores are zero in the first sitting. As a consequence the sum of all the ind ces in / J cannot be divided by the same number of subjects as can the sum of the scores in sittings one and two. Thus, in Table 16, the sums in sittings 1 and 2 for the boys are divided by 16 and 15, respectively, while ths sum of the indices in / f is divided by only 6. The zero scores in 1 and 2 cannot be represented in / ';, except when the zero is in the second sitting. The percentage of improvement in the second sitting cannot be ascertained when the initial score is zero, but when the score in the second sitting is zero there is an absolute loss which can be expressed as .00. This difficulty is frequently complicated by occasional absences, which makes it impossible to ascertain the index or makes it necessary to base the index on another sitting, so that frequently the data averaged in the index column are not identical with the data in the score columns. Under these conditions it is inevitable that discrepancies should arise. When there are no zero scores in the different sittings and when the subjects are all present the results by the two methods should harmonize. We base our conclusions regarding the Normal and Epileptic School Children 21 improvement made in the test largely on the indices. The reader can base the comparison on the first method, should he so desire, from the data sup- plied in the tables. The Ave" in the last column was obtained by finding the difference be- 1-5 tween the percentage scores in each successive sitting, 1 to 5 (the scores in the columns numbered 1, 2, 3, 4 and 5), then substracting the sum of the losses from the sum of the gains (or vice versa when the sum of the losses is larger than the sum of the gains) and dividing the remainder by the number of differences. To illustrate: the successive scores for subject 80 in Table 5, are 23.3. 16.6. 30, 13.3 and 50. The successive differences between these scores are -6.7, +13.4, -16.7 and +36.7. The sum of the gains ( + ) is 50.1 and of the losses (-) 23.4. Hence the total net gain for the 4 last sit- tings is 26.7, and the average gain for each successive sitting (i. e., the last four sittings) is 6.7. For subject 82, the successive scores are 36.6, 36.6, 33.3, 33.3 and 33.3. The successive differences are 0, -3.3, 0 and 0. The total loss for the 4 last sittings is 3.3 and the average loss .8. The averages in the tables for "B", "G", and "B & G" at the bottom of the columns containing the percentage scores, were obtained by finding the sum of the individual scores for all the subjects and dividing by the total number of subjects in a given column. These averages were not obtained by averaging the different averages in the per cent columns. The final aver- ages (at the bottom) for the columns containing the indices of improvement (j) were obtained in the same manner. 1 he computation of a large amount of the data in the tables was en- trusted to two paid and two or three volunteer assistants. So many errors, however, were discovered that it was necessary to revise practically all of the tabular data. Doubtless occasional undiscovered errors in mathematics still remain, but it is not believed that these are numerous or serious enough to exert any appreciable effect on the final results. In the following chapters we shall (1) reproduce or describe the test materials, (2) state the nature or purpose of the tests, (3) give the directions for administering the tests, (4) present the tabular data, (5) state some of the more important empirical results, and (6) finally summarize a few of the conclusions. We would emphasize that our conclusions are based upon the groups which we have actually studied and, owing to the limitations of these groups, must be regarded as more or less tentative. One chapter will be de- voted to each test. In a final chapter we shall give a brief summary of con- clusions. The descriptions of the tests will be somewhat brief in view of the fact that detailed directions for administering the tests will be given. The sections dealing with the description and with the administration of the tests are intended to supplement one another. II. MEMORIZING VISUALLY PRESENTED NUMBERS Immediate and Deferred Recall of Three-Place Digits: Al to AS 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS STIMULUS MATERIALS The following figures were printed on cardboards1 7 inches wide by 2G inches long. The size of the type was such (1% inches high) that the figures could be read easily from the most remote seat occupied by anyone taking part in the tests. Each card contains ten three-place digits. 961 057 578 149 462 666 723 103 192 294 999 730 555 275 429 812 634 815 306 786 281 037 340 954 480 821 043 678 518 307 Al A2 A3 185 590 888 714 238 305 098 111 • 219 026 983 763 479 409 076 201 726 934 333 875 152 857 064 690 560 657 421 642 342 547 A4 A5 A6 RECORD SHEETS The pupils' records were made on plain sheets of paper \x/-> by 7% inches, numbered from Al to A5. These blank sheets were prepared by hand and are not included in the set published by the C. H. Stoelting Co. 'The completes t of test materials will be printed although, as already explained, the sixth series was not given. 22 Normal and Epileptic School Children 23 In the arrangement of the figures on the different cardboards care was taken that no number combinations would recur on the same card or on the successive cards, with the two following exceptions: First, the digits were made identical in one of the ten three-place figures on each card, thus: (1) 555. (2) 999. (3) 666, (4) 333, (5) 111, and (6) 888. These figures were given a different position on each card, so that the subjects would not acquire the habit of looking for a figure with identical digits in the same position in the series. It is possible that the subjects began to anticipate a figure with identical digits after the second or third sitting, although our results do not clearly indicate this, as will be seen later. Second, one figure in each card began with a cipher, thus: (1) 037, (2) 057, (3) 043, (4) 098, (5) 064, and (6) 076. These figures were also given a different position in the series on each card. The figures were systematically distributed on each card in such a way that each of the ten digits would occur but once in each vertical column. It was believed that a systematic distribution of the numbers in accordance with this plan would be most likely to yield a series of numbered cardboards of equal difficulty. Xo empirical tests, however, were carried out to demonstrate whether the successive number cards were uniform in diffi- culty. To make such a determination it would be necessary to give one each of the different number cards to different groups of children of the same intelligence status. The series of cardboards, however, can be validly used to measure the comparative performance of normal and epileptic children, even though they should prove not to be exactly equal in difficulty. 2. NATURE OR PURPOSE OF THE TEST This test aims to measure the visual memory span for non-sense material (three-place digits). Or perhaps it would be more accurate to say that the test measures the ability to reproduce memorized non-sense materials, for it should be emphasized that this test requires actual memorizing during a period of 45 seconds, unlike the tests of memory (or attention) span proper, in which a single impression is given through a momentary exposure or a single repetition. The strength of memory (memory span) in the latter cast- depends upon the primary impressionability and retentiveness of the neurones, while the strength of memory in the former case depends upon these factors plus the depth of the impression produced by repetition, ^'e have measured the subject's retentiveness of the memorized material in this test both by the method of immediate recall and of deferred recall. 3. DIRECTIONS FOR GIVING THE TESTS General Instructions: follow the general directions on pages 15 and 16. Order of Giving Tests: first sitting, Al; second sitting, A2; third sitting, A3; fourth sitting. A4; etc. 24 The Measurement of Mental Traits in Time: Display the cards for exactly forty-five seconds. Allow the children exactly sixty (60) seconds in which to write from memory. Hold the cardboards so that the number can be easily seen by all the pupils in the room. Instructions to be Given to the Pupils: "On the other side I have a set of numbers. The moment I turn the card around I want you to start to learn (memorize, commit) just as many numbers as you can. I want you to study just as hard as you can, because I am only going to let you see them for a little while. Then when I turn the card around again. I want you to turn your papers over at once (this, however, is not necessary with these vacant blanks. But this method was followed in giving the tests in this experiment, in order to employ uniform conditions in all the tests), and write down just as many numbers as you can remember. Write them just in the order you saw them, if you can; the first number at the top, the second num- ber next, and so on. But if you cannot remember the order, write them as well as you can. (The important thing, of course, is to have the three digits in each number in the right order.) When I say 'stop' you must stop at once and turn your papers over again." "Now, ready!" "Turn!" "Now, stop!" "Turn your papers over." At the Close of the Experiment: Pupils write full name, age, sex, date, hour, condition, school, and grade. Ranking — How to Mark the Papers: Scale of 100. Each correct number (all three digits right and correctly placed), 10%. One digit of a number placed correctly, 3.3%. Two digits placed correctly, 6.6%. The sum of two of these should be scored 10%. Always give results in terms of the per cent of efficiency (not ineffici- ency). Xo deductions have been made in this experiment for incorrect numbers. On this rather liberal basis of crediting care must be taken lest credit be given for mere guesses. Combinations, if partly right in two or three figures (three-place groups), must be credited only once. Special records may be kept of 555, 999, 666, 333, 111, 888, and of 037, 057, 043, 098, 064, and 076. Deferred Reproductions: At the beginning of the second sitting the ex- perimenter may, if he chooses, ask the pupils to reproduce Al, without a sec- ond exposure, and similarly during any later sitting. But the pupils must not be told that they will be tested again on the test at the next sitting. Normal and Epileptic School Children 25 TABLE 4 Immediate Reproduction of Memorized Three-Place Digits: Al— A5 Normal Children CiO. Ae. bx. 1 % * -. • '*) % f ■ t f\ 4 V, ft 5 •A f \ 1-2 V< 4-5 4-3 i-t fi 1 - - Gain l-5 2 7 B 31 6 25.0 1 10 29 9 1 36 35 0 1 19 20 0 .55 28 3 27 5 1 31 1 02 — 28 37 2 .7 G 28 3 30 0 i 28 28 3 90 44 9 1 62 33 3 74 27 4 39. 1 1 42 1 18 33 3 14 4 7 Av. 27 4 26 3 i 16 29 1 1 13 39 9 1 87 26 6 !$4 27 8 33 3 1 36 1 10 30 g —1.3 g g B ' 23 3 29 5 i 32 33 3 1 '.2 37 1 1 09 2" 5 86 26 4 32 6 1 19 1 29 29 1 ^5 3 g G 17 7 32. \ \ 97 32 1 1 i 2 36 6 1 09 27 7 88 24 9 32 1 1 35 1 £8 29 2 2 4 11 g Av. 21 8 30 2 \ 49 32 9 1 09 36 9 1 09 29 0 86 26 0 32 5 1 26 1 45 29 6 1 .8 3 9 B 17 7 45 0 i 75 44 4 2 19 46 6 1 0b 47 7 1 !02 31 6 47.1 1 gg 361 39 6 8T 4 9 G 19 9 26 6 i 49 35 5 5 23 26 6 85 25 8 1 10 24 1 26 2 1 15 1 36 26 3 L6 7 9 Av 19 0 33 9 i 99 39 9 1 71 35 2 94 35 2 1 06 27 7 35 2 1 ^32 2 32 32 0 4 5 4 10 B 14 9 50 0 i 04 49 9 98 52 5 1 15 51 6 1 02 45 6 52 0 1. 14 1 15 48 2.6 3 10 G 31 1 48 8 i 62 51 5 1 03 46 6 97 37 7 '80 39 9 42 1 1 13 1 38 42 6 16 7 10 Av. 39 0 49 4 i 33 50 7 1 CO 49*9 1 07 45 7 93 43 3 47 .8 1 14 1 25 46 0 2 1 3 11 B 39 2 32.2 i 01 0 1 1 Q 45 5 i i £o aq >.> O 75 1'? ■j- i i 39.4 1 19 1 02 35 q 1 1 8 11 G 38 7 41.6 i 12 45 g i — > 0/ 0 .98 Ml WJ i L 42 4 1.06 1 04 4 1 1 7 11 11 Av. 36 9 390 i 09 39 0 1 04 45 7 1 23 3ft 3 91 37 9 41 6 1 09 1 03 39 7 8 5 12 B 41 4 44.6 i 34 43 3 1 00 48 3 99 51 9 1 24 43 0 48 6 1 '. 13 1 41 45 1 2 7 3 12 G 39 9 44 .4 i 28 38 8 87 53 3 1 41 51 1 98 42 1 52 2 1 26 1 64 45 4 2 8 8 12 Av. 40 8 44 .5 l 32 41 6 95 50 4 1 17 51 6 U4 42 6 49^9 1 18 L50 45 2 2.7 4 13 B 39 1 48.8 l 23 54 9 1 07 49 9 89 43 3 94 15 7 46.6 1 01 1.24 47 1 1.1 2 13 G 48 3 43.3 84 46 6 1 23 56 6 1 27 51 6 90 45 7 54.1 1 21 1.03 49 2 .8 6 13 Av. 42 2 46.6 i 07 52 2 1 13 52.1 1 02 46 0 93 45 7 49.1 1.08 1.17 17 8 1.0 4 14 B 43 2 49.9 l 43 14 9 91 51 0 1 06 36 6 1 11 46 5 44.7 1.01 1 45 46 2 1.2 7 14 G 42 8 47 1 i 15 45 3 1 (9 42 8 82 50 4 1 3g 44 9 46 6 1 06 1 28 46 9 2 0 11 14 Av. 42 9 48.1 l 25 45 1 1 01 45.3 89 46 3 L29 45 5 47.6 1.07 1 33 43 3 1.7 2 15 B 55 0 70.0 i 32 76 6 1 10 66 6 .87 62 5 71 6 1 17 1.29 67 0 3 8 2 15 G 43 3 51.6 i 20 48 3 93 65.0 1 34 48 3 74 47 4 56.6 1 19 1 12 51 2 1.2 4 m Av. 49 1 60.8 i 26 48 3 93 70.8 1 22 57 4 .80 54 9 64.1 1 18 1.20 59 1 2 3 1 16 H 56 6 46.6 82 40.0 86 50 0 1.25 51 6 45.0 .87 .88 48 3 —2.2 3 16 G 41 .1 36.6 i .18 53 3 i 31 49.9 93 51 0 1.02 42 7 50.5 1.56 1.37 47 5 4.6 4 16 Av. 44 9 39.9 i 06 53 3 l 31 47 4 91 50 8 1 08 41 9 49 1 1.38 1.25 47 7 3 0 1 17 B 66 6 46.6 .69 60 o l 28 73.3 1 22 63 3 .86 56 6 68.3 1.20 .95 61 9 - .8 2 17 G 43 3 36.6 .78 38 3 i 00 48.3 1 26 48 3 .99 40 8 48.2 1 18 1.12 43 8 1 4 3 17 Av. 51 0 41.6 73 45 5 l 09 56 6 1 24 53 3 .94 46 0 54 9 1 18 1 06 49 8 .6 Ave. B 36 4 41.9 l 31 42 2 l 18 47 9 1 07 42 1 96 39 0 44 9 1.16 1.43 42 0 1.8 Ave. G 36 0 41.4 l 27 41 2 l (5 45 1 1 11 41 7 1.01 38 5 43.4 1.19 1.30 41 2 1.8 Ave. B&G 36 2 41.6 l 29 42 0 l 11 46.4 1 09 41 9 99 38 8 44 0 1.18 1.36 41 6 1.8 26 The" Measurement of Mental Traits in TABLE 5 Immediate Reproduction of Memorized Three-Place Digits: Al— A5 Epileptics Su. Ae. Sx. I % i % J l s /! * ft 5 % f\ 1-2 % 4- fi 5 m f\ Is : * 5 Gain Ave. %. 81 10 B 16.6 26 6 1 60 16.6 62 16 6 2 1 6 i 30 I 00 19 9 0 102 10 G 39 9 40.0 1 . 02 26 6 66 30 0 ] 12 30.0 1 00 39 9 30 0 75 75 33 3 -2 4 104 10 G 13 2 6.6 .50 23.3 3 53 20 o 85 26.6 1 33 9 9 23 3 2 35 2 01 17 9 3.3 Av. G 26.5 23.3 .76 24 .9 2 09 25 0 98 28.3 1 16 24 9 26 6 1 55 1 38 25 6 .4 99 11 G 39.8 16.6 .41 48.6 2 80 20.0 43 28 2 33 3 1 18 50 30 7 -6.6 79 12 H 13.3 30.0 2 . 25 36.6 1 22 23 3 63 33 3 1 42 21 g 28 3 1 31 2 50 27 3 5.0 P4 12 B 23.3 43.3 1 .85 30.0 69 33 3 1 11 23.3 69 g; 3 28 3 84 1 00 30 6 0 91 12 B 26 6 40.0 1 .50 36.6 91 43 3 1 18 50.0 1 15 33 3 46 6 1 39 J 87 39 3 5.8 Av. B 21 0 37 7 2.80 34 4 94 33 3 97 35.5 1 08 29 4 34 4 1 18 1 79 32 4 3.6 108 12 G 16.6 6.6 39 20 0 3 03 20 o 1 00 23.3 1 16 11 f, 21 6 I 86 1 40 17 3 1 6 93 13 B 20.0 36.6 1 .83 13.3 36 23 3 1 75 26.6 1 14 28 3 21 9 87 1 33 23 9 1 . 6 86 13 B 26.6 16.6 62 29.9 1 8D 33 3 1 j J 10.0 30 21 g 21 6 1 00 37 23 2 -4 1 Av. B 23.3 26.6 1 .22 21 .6 1 08 28 3 i 13 18.3 72 24 9 23 2 93 85 23 5 -1 2 97 13 G 23. 1 20 0 .86 20.0 1 I 0 30.0 1 50 21 5 25 0 1 16 1 29 23 2 2.3 77 14 B 33.2 20.0 60 30.0 1 50 30 0 1 00 26 6 88 26 g 28 3 1 06 80 27 9 -1.4 89 14 B 43.3 40 0 .92 43 .2 1 08 46 6 1 07 23 .3 50 41 6 34 9 83 53 39 2 D U Av B 38 2 30 0 .76 36.6 1 29 38 3 1 03 24.9 69 34 1 31 6 94 66 33 5 -3.3 80 15 B 23 3 16.6 71 30.0 1 80 13 3 44 50 0 3 75 19 9 31 6 1 58 2 14 26 6 6 7 82 15 B 36.6 36.6 2 .00 33.3 90 33 3 I 00 3'!. 3 1 00 36 fi 3:1 3 9i i 90 34 6 - 8 83 15 B 3 .0 20.0 .66 23 3 1 16 10 0 42 26 6 2 66 25 0 IS 3 73 88 21 9 - 8 Av. B 29.9 24.0 79 28 8 1 28 18 S 62 36.6 2 47 27 1 27 7 1 07 1 30 27 7 1 - 6 103 15 G 20.0 0. .0 13.3 36 6 2 75 30 0 82 10 0 33 3 3 33 1 50 19 9 2.5 78 16 B 10 0 13.3 1.33 20.0 1 50 30 0 1 45 30.0 1 00 11 6 30 0 2 61 3 00 20 6 5.0 88 16 B 20.0 40 0 2'0 40.0 1 00 40 0 1 00 30 0 75 30 0 35 0 1 16 1 50 34 0 2.5 85 16 B 20.0 36.6 1.83 30.0 81 53 3 1 77 30 0 56 28 3 41 6 1 46 1 50 33 9 2 5 Av. B 16 6 29.9 I 72 30.0 1 10 41 1 1 40 30.0 77 23 3 35 5 1 74 2 00 29 5 3.3 96 16 G 10.0 10.0 1 00 20.0 2 00 10 0 50 16.6 1 66 10 0 13 3 1 33 1 66 13 3 1.6 98 16 G 40.2 46 6 1 15 36 6 78 56.6 1 54 43 4 46 6 1 15 1 40 45 0 5.4 106 16 G 23 3 23.3 1.00 23 3 1 00 20 0 85 33.3 1 06 23 3 26 6 1 14 1 42 24 6 2.5 100 16 G 9.9 33.3 3.36 33 3 1 00 26 6 79 40.0 1 50 21 6 33 3 1 54 4 04 28 6 7.5 Av. G 20.8 22.2 1.78 30.8 1 28 23 3 73 36.6 1 44 21 4 29 9 1 29 2 13 27 8 4.2 87 17 B 46 6 26.6 57 43.2 1 62 10 0 23 30.0 3 00 36 6 20 0 54 64 31 2 -4.1 90 17 B 10. n 0 M 30.0 20 0 66 30.0 1 50 5 0 25 0 5 00 3 00 18 0 5.0 Av B 28.3 13.3 .28 36.6 1 62 15 0 44 30.0 2 25 20 8 22 5 2 77 1 82 24 6 4 95 17 G 0. 0 .0 0. 10 0 0. 0 0 5 0 2 0 0 105 18 G 23.2 20 0 .86 30.0 1 45 20 0 66 23.3 1 16 21 6 21 6 1 00 1 00 23 3 .02 107 21 G 29.8 36 6 1.22 23.3 63 20 0 85 43.3 2 16 33 2 31 6 95 1 45 30 6 3 3 101 24 G 13.3 6 6 .49 20. 0 3 03 16 6 83 26.6 1 60 9 9 21 6 2 18 2 00 16 6 3.3 Ave. B 24.9 27.7 1 17 31 2 1 .16 29 3 1 .02 29.3 1 •0 26 3 29 3 1 41 1 43 28 5 1.1 Ave. G 21.5 16.9 .92 24.7 1 .77 22 2 .99 28 5 1 20 19 3 25 6 1 54 1 57 22 9 1.8 Ave B&G. 23.3 22.7 1 06 28.2 1 44 26 2 1 .01 28.9 1 26 23 0 27 6 1 46 1 49 25 9 1.4 For explanation of all the tables see pages 17 to 21. 5. RESULTS (See Tables 4 and 5) IMMEDIATE RECALL Most of our comparisons in this and the succeeding sections will be based on the figures for the two sexes at the bottom of the tables, which give the general averages for all ages, and on the figures in the next to the last column ('" '), which give the averages for all five sittings. The figures used in discussing the comparative progress will be based on these averages, unless otherwise stated, which yield the most reliable figures since they are based on all the subjects in each group and on all the sittings. Normal and Epileptic School Children 27 comparative efficiency The average efficiency per subject based on all the five tests for all the normals1 was 41.6%, and for the epileptics 25.9%. The epileptic group showed only 62% of normal capacity. The average for the epileptics is less than the average for the youngest normal children who were tested. The average, based on all the five sittings and all the members of the groups, is higher for the boys than for the girls among both normals and epileptics, but the difference is very small among the normals. The girls' capacity amounts to 98% of the boys among the normals as against 80% among the epileptics.2 The boys' superiority, based on the averages for the entire group, is manifest in all the sittings among both the normals and the epileptics. Based on the general averages for the five sittings, the normal boys are superior to the normal girls in five ages, inferior in five, while the averages are equal in one age. Among the epileptics the boys are superior to the girls in 5 ages, inferior in one age, while no comparisons are possible in the remaining ages. It should be remembered, however, that comparisons between the two sexes in the different ages, in all the tests, must be merely tentative, because of the fewness of the subjects in the different ages. The performance in this test among the normals tends to improve with increasing age. but the improvement is subject to considerable fluctuation. Thus while the scores in the two highest ages (47.7% and 49.8%) are em- phatically higher than in the two lowest ages (30.8% and 29.6%, averages for the five sittings), there is an increase from age to age in only six ages, while in four ages there is an actual loss (comparing the score in a given age with the score in the next preceding age). These losses can be accounted for — we shall see that they occur in the other tests also — by two circum- stances. First, the limited number of subjects tested in each age would necessarily cause irregularities in the age curve. Second, a preponderance of dull or average pupils or a lack of bright pupils in any age would tend to lower the score in that age. As already stated, we tried to secure an equal number of dull, average and bright pupils in each grade, but did not entirely succeed in this attempt. Worse still, some pupils dropped out or had to be dropped, and this created a still greater disproportion in the relative number of bright, average and dull pupils in some grades. Moreover, the pupils in the different grades were not always of the same age. When rearranged into age groups inequalities likewise arose in the number of dull, average and bright pupils in each age. The actual distribution of dull, average and bright pupils in each age is given in Table 5a. As we have already explained, the classification of the pupils was made by the school staff. In this test the losses occur in ages 8, 11, 14, and 16, while the score is also low in age 17. It will be observed that 72% of the 8-year-olds are classi- 'For the sake of brevity we shall use the word "normals" for the normal children. -The figures on which these and similar comparisons are based may be found in the tables. 28 The Measurembnt of Mental Traits in TABLE 5a Number of Dull, Average and Bright Pupils among the Normal Children in each Chronological Age. Boys Age 7— Dull . . . Average Bright Total . , Age 8— Dull Average . . . Bright Total . . Age 9- Dull Average Bright Total . Age 10— Dull Average. . Bright Total . Age 11— Dull Average . . . Bright .... Total Age 12— Dull Average . . . Bright .... Total . Age 13— Dull Average. . . Bright Total Age 14— Dull Average . . Bright Total . Age 15 — Dull Average . . . Bright Total . . Age 16— Dull Average . . Bright Total Age 1 7— Dull Average. . Bright . . Total 1 1 0 2 1 5 2 <* 1 1 1 3 i 2 2 4 2 1 0 3 1 2 2 n 2 2 4 2 0 2 4 0 1 1 2 1 n o l o l o l Girls Normal and Epileptic School Children 29 fied as dull or average, compared with 50% for the 7-year-olds. Only ")?' , of the 10-year-olds are classified as dull and average (most of these being average), while 72% of the 11-year-olds are so classified (half of these being dull. In age 14, 63% were classified as average and dull, most of these being dull, while only one-half of the 13-year-olds were classed as aver- age and none as dull. In age 16, 75% — and in age 17, 100% — were classed as dull and average, while in age 15 no one was classed as dull, while 75% were classed as bright. The best record was made by the 15-year-olds. From these figures it would seem that the plateaus or depressions in the age curves are accounted for, in part at least, by a preponderance of dull or aver- age children, or a lack of bright children in the ages which show losses. As was to be expected, there is no consistent gain with increasing chronological age among the epileptics. The scores are about the same for the older as for the younger epileptics. Had the data for the epileptics been classified according to mental (Binet-Simon) instead of chronological age, we would probably have found an increasing degree of efficiency with ascend- ing Binet-Simon age. Owing to the lack of time we did not find it possible to retabulate the data according to mental age. This can be done, however, from the data supplied in the various tables. Table 2 gives the Binet-Simon age for each epileptic. The highest score is made by the normals in the fourth sitting and by the epileptics in the fifth, and the lowest score is made in the first sitting by the normals and in the second sitting by the epileptics. Xot a single pupil among the normals failed completely in any sitting. Two epileptics, Xos. 90 and 103, failed in one sitting each, namely in the second, while one epileptic, Xo. 95, made no score in four sittings. Xo. 103 had had a convulsion 10 minutes before the test in which she failed, and made no effort to do anything, while Xo. 90 was stupid in the second sitting, and made little effort, particularly in E. G and J. Xo. 95 had two convulsions during the night preceding the second sitting. She suffers from post-con- vulsive stupor, and made little effort. COMPARATIVE IMPROVEMENT The average monthly gain (the final column in the tables) for each of the normal subjects amounted, in absolute units, to only 1.8% and for the epileptics only 1.4%. The epileptics improved 77% as much as the normals. It should be remembered that this figure is based upon absolute units. Com- paratively, the epileptics did better than this since the absolute size of the score was smaller for the epileptics than for the normals. This may be shown by basing the improvement on the indices (/). When the comparison is based on the indices of improvement, which show the percentage of im- provement made by the normals in terms of their own scores made in earlier 30 The Measurement of Mental Traits in sittings, and similarly for the epileptics, the results are as follows: The gain in the fifth sitting compared with the first (/{) amounted to 36% for the normals and 49% for the epileptics, a difference of 13% in favor of the epileptics. The gain, based on the comparison of the average for the last two sittings with the average for the first two ( / ) amounts to 18% for the normals and 46% for the epileptics, a difference of 28% in favor of the epileptics. The following are the percentages of improvement made in each of the sittings when the records are compared with the scores made in the next preceding sitting: Sitting 2d 3d 4th 5th Normals 29% 11% 9% -1% Epileptics 6% 44% 1% 26% The epileptics' gain was 23 '/< greater than the normals' in the second sitting, 33/i greater in the third sitting, 27' ; greater in the fifth and 8% less in the fourth. The greatest improvement by the normals was made in the second sitting, and by the epileptics in the third sitting. The relative amount of gain made by the normals steadily decreased from sitting to sitting. In fact, there was a slight loss in the fifth sitting compared with the fourth. On the other hand, the gains made by the epileptics were very ir- regular, very large gains being made in the third and fifth sittings and only small gains in the second and fourth sittings. An examination of the individ- ual records in Table 5 will show that the fluctuation is large for a consider- able number of the epileptics. This was to be expected. The mental capacity of epileptics subject to periodical attacks varies greatly from day to day. On some days they can accomplish a considerable amount of work, while on other days they can do very little. Based on the average monthly gain for all the sittings (last column), eight of the epileptics lost while three made no gain, or 36% of all the epileptics, while the corresponding figures for the normals are 20 and 6, or 34' ; of all the normals. (As we have already explained the individual records for the normals are not published here.) The absolute average monthly gain for the normal boys was just the same as for the normal girls, but it was 61% greater for the epileptic girls than for the epileptic boys (1.8% vs. 1.1%). Based on the indices of im- provement, the normal boys gained more than the normal girls in three of the indices, and less in three, while the same result obtains among the epileptics. The gain in the fifth sitting compared with the first (/ \) was 13% greater for the normal boys than for the normal girls (43% vs. 2,0' < ). while the gain for the epileptic girls was 14' "< greater than for the epileptic boys (57', vs. 43' ( ). The gain based on the last two sittings compared with the first two (/ f-i) was 3% greater for ths normal girls than for the normal boys (19% Normal and Epileptic School Children 31 vs. 16%) and 13% greater for the epileptic girls than for the epileptic boys (54% vs. 41%). The comparative data for the boys and girls of different ages can be obtained from the tables. Based on the average gain in all five tests, last column, and the index of there is no clearly defined relation between the amount of the improve- ment and the age of the children. In half of the ages for the normal children the improvement grows larger with ascending age, while in half of the ages it grows smaller, when the score in each age is compared with the score in the next preceding age. If any correlation exists it is obscured by the limited number of cases in each age. We should not expect to find any constant correlation among the epileptics, and none is found, owing to the varying degrees of mental arrest found among these subjects of different ages. SUCCESS with figures containing identical digits and initial ciphers We have made a separate tabulation in Tables 6 and 7 of the results for the figures in which all the digits were the same and for the figures which began with a zero, each card containing one each of these figures. It will be somewhat difficult to compare directly the scores made in Tables 6 and 7 with the scores made in Tables 4 and 5. because the scores made in the latter tables are based on all the ten figures, including those with identical digits and initial ciphers, while the scores in each column of Tables 6 and 7 are based only on one figure. TABLE 6 Percentage of Pupils having Entirely Correct < all three Digits Correct) the following Special Combinations in the Digits Memory Test: Al— 5. Normal Children Grade January February March April M ay 555 037 999 057 666 043 333 098 111 064 % cfo % fo °h % fo 1o % % \\i 33-3 ~v 66.6 50. 83.3 0 66.6 0 100. 16.6 87.5 37.5 62.5 25. 75. 50. 25. 87.5 87.5 25. 88.8 22.2 100. 55.5 100. 55.5 100. 55.5 66.6 44.4 90.9 54.5 81 5 72.7 81.8 36.3 72.7 54 .5 100. 88.8 45.4 66.6 77.7 88.8 88.8 100. 66.6 100. 100. 77.7 VII . 75. 66.6 91.6 91.6 66.6 50. 83.3 83.3 91.6 58.3 71.4 42.8 57 1 42.8 100. 28.5 57.1 71.4 100. 71.4 75. 75. 75. 75. 50. 50. 75. 75. 100. 50. 100. 50. 100. 75. 75. 25. 75. 75. 100. 75. 83.3 83.3 83 3 66.6 50. 0 100. 83.3 100. 66 6 77.6 51.3 81.5 65.7 80.2 39.4 76.3 69.7 92.1 52.6 32 The Measurement of Mental Traits in i TABLE 7 Percentage of Pupils having Entirely Correct (All three Digits Correct) the following Special Combinations in the Digits Memory Test: Al— 5. Epileptics B.-S. Age January February March April M ay 555 °h 037 * ■ 999 % 057 % 666 % 043 9i> 333 % 098 % 111 0i6 20. 0 20. 0 60. 0 40. 20. CO 0 9 20 20. 40. 40. 100 40 26 40. 60 40. 10 47 11.8 41.1 23 5 88.2 11 8 35.3 35 3 100 11 8 100 0 100 66 100 100 100. 33 3 100. 33 3 43.3 10. 43.3 26.6 86.6 23.3 40. 33.3 86.6 16.6 We have computed the results in Tables 6 and 7 in terms of the per- centage of pupils in each grade or each Binet-Simon age who made perfect reproductions, i. e., had all three digits of each figure correct. These per- centages, however, can be readily converted into percentage scores which are directly comparable with the percentage scores in Tables 4 and 5. Thus the score 36.2, which is the general average for the first sitting in Table 4, indi- cates that the pupils scored 36.2% of a possible score of 100', ; , since the score is based on 10 figures for the complete reproduction of each of which 10% vvas allowed. On the other hr-nd, the score 77.6'i at the bottom of the 555 column in Table 6, indicates that 77.6' , of the pupils reproduced the figure correctly and, accordingly, that the average sgore was 7.76%, for if we divide 77.6% by 10 (since the results are based on only one figure of the value of 10'4 ) we obtain 7.76% as the average score for 555. But 7.76% is 77.6% of 10% , which represents a perfect score for one figure. We shall note only a few of the more important results. The scores for the figures in which all the digits were the same (Tables 6 and 7) are emphatically higher in every case among both the normals and the epileptics than the scores based on all the figures in the corresponding sittings. Thus, the average score for 555 is 2.1 times higher for the normals and 1.8 times higher for the epileptics than the average scores for the normals and epileptics, respectively, in sitting 1 of Tables 4 and 5; for 999 the figures are 1.9 higher for both the normals and the epileptics than the average scores in sitting 2; for 666, 1.9 and 3. higher, respectively, than the scores in sitting 3; for 333, 1.6 and 1.5 higher, respectively, than in sitting 4; for 111, 2.2 and 2.9 higher, respectively, than the scores in sitting 5. When the scores are based only on the figures with identical digits the normals did relatively somewhat better than the epileptics (when compared, of course, with the records in Tables 4 and 5) in the first and fourth sittings, the epileptics did noticeably better than the normals in the third and fifth Normal and Epileptic School Children 33 sittings, while there is no difference in the second sitting. The epileptics on the whole do better relatively to the normals when rated by these figures than when rated by all the figures. In fact, the absolute score for 666 was 6.4'J higher for the epileptics than for the normals. In all the other sit- tings, however, the normals did better, the difference for 555 amounting to 34.3'/, ; for 999. 38.2£ ; for 333. 36.3',; ; and for 111. 5.5%. The separate tabulation of the results for these figures — and the figures for the initial ciphers — was made primarily to determine whether the im- provement from sitting to sitting might not be explained by a possible tendency on the part of the subjects after the second trial to expect these figures and thus make a special effort to impress them on their memories. But this does not appear clearly to be the case. The normals made their highest score for the identical digits in the fifth sitting, and the lowest score in the fourth sitting, while they made their highest score in Table 4 in the fourth sitting and the lowest score in the first sitting. The epileptics made their highest scores for the identical digits in the third and fifth sittings and the lowest score in the fourth sitting, while in Table 5 the highest score was made in the fifth sitting and the lowest in the second. When we turn to the figures with the initial cipher we find a somewhat different situation. The average scores are higher than the corresponding average scores for all the figures in Tables 4 and 5 in four sittings for the normals and two for the epileptics, while they are lower for the normals in one sitting and for the epileptics in three sittings. The average score for 037 is 1.4 times higher for the normals, but less than one-half as large for the epileptics as the average scores for all the figures for the normals and epil- eptics, respectively, in sitting 1 of Tables 4 and 5; for 057 the figures are 1.5 and 1.1 higher for the normals and epileptics, respectively, than for all the figures in sitting 2; for 043 they are .93 and .82. as large for the normals and epileptics, respectively, as for all the figures in sitting 3; for 098 they are 1.5 and 1.2 higher, respectively, than the per cents for all the figures in sitting 4; and for 064 they are 1.2 higher for the normals and a little over one-half as large for the epileptics as the per cents for all the figures in sitting 5. It was thought that the pupils would be more likely to retain these figures than any ethers except those with identical digits, because of the unusual character of the initial digit (factor of primacy), but this appears not to have been the case except for the normal pupils. The epileptics were less aided by this arrangement than the normals. It is evident, therefore, when we compare the scores in Tables 4 and 5 with the scores in Tables 6 and 7. that there is no evidence that the average gain from sitting to sitting (based on the ten figures) was increased by a growing tendency to pay special attention to or specially to memorize the identical figures or the figures containing initial ciphers. M T 3 34 The Measurement of Mental Traits in PERCENTAGE OF EFFICIENCY BASED ON THE FIGURES IN WHICH ALL THREE OF THE DIGITS WERE CORRECTLY REPRODUCED In view of the fact that our scores are the result of summating total and partial credits — total credits being given when all three digits were correctly recalled and partial credits when only one or two digits of a figure were cor- rectly reproduced — it seems desirable to ascertain how large the scores are when the scoring is based only on the figures which were recalled entirely correct. Tables 8 and 9 give the average percentages of efficiency when the scoring is based only on the figures in which all three digits were correctly recalled. TABLE 8 Percentage of Efficiency when based only on the Total Number of Figures in which all Three Digits were Recalled Correctly. Normal Children January February March April May Grade Tot.' Tot.i Tot.i Tot.' Tot.i II ... 4 6.6 9 15 0 8 16.0 10 16.6 13 23.3 Ill 13 16.2 9 15 0 18 22.5 20 25.0 18 22.5 iv ....... 20 22.2 26 28.8 31 34 4 32 35 5 28 31.1 28 25 4 30 30.0 29 29.0 36 36 0 37 33.6 VI 28 31 1 34 42 5 39 43 3 36 45.0 29 41.4 VII • .. 41 34.1 47 39.1 34 30.9 51 46.3 44 40.0 24 34.2 15 30 0 30 42.8 26 37.1 26 37 1 l irst High 11 27.5 11 27.5 12 30.0 14 35.0 13 32.5 Second High 15 37.5 20 50.0 9 22.5 22. 55.0 20 50.0 Third High 22 36.6 21 42.4 11 27.5 28 46.6 26 43.3 Total 206 27.1 222 31.7 221 31.1 275 37.6 254 34.7 'Total number of figures in which all three digits were correctly recalled. -Average per cent of efficiency. Normal and Epileptic School Children 35 TABLE 9 Percentage of Efficiency when based on the Total Number of Figures in which all Three Digits were correctly Recalled. Epileptics c . . . oUujC'Ct Age Sex January Total fo February Total % March Total % April Total 1h May T tal % . 81 10 B 1 2 1 102 10 G 3 3 2 3 2 104 10 G 0 0 2 2 2 99 11 G 2 0 4 1 79 12 B 1 3 3 2 2 84 12 B 1 4 .3 2 2 91 12 B 4 3 4 4 108 12 G i" 0 2 1 2 " 93 13 B o 1 1 2 1 86 13 B 1 2 3 1 97 13 G 1 2 1 77 14 B 1 2 2 2 2 89 14 B 4 3 3 3 1 80 15 B 2 1 3 1 5 82 15 B 2 3 3 3 2 83 15 B 3 2 2 1 2 103 15 G 2 0 0 3 2 78 16 B 0 1 1 3 3 88 16 B 1 4 4 4 3 85 16 B 0 2 2 4 2 96 16 G 1 1 2 I 1 98 16 ' G 3 0 4 2 4 106 16 G 9 2 2 1 2 100 16 G 0 2 3 0 2 87 17 B i 2 3 1 2 90 17 B 1 0 3 2 3 95 17 G 0 0 0 1 0 105 18 G 1 0 2 2 2 107 21 G 1 3 1 2 3 101 24 G 1 0 1 1 2 Total B 21 15.0 33 22 0 38 25 3 39 24.3 36 22 5 Total G 18 12.8 13 9.2 26 18.5 19 15.8 25 19.2 Total ..B&G 39 13 9 46 15.8 64 22.0 58 20.7 61 21.0 It is obvious from these tables that most of the credits made by both the normals and the epileptics were earned from the figures all three of whose digits were correctly reproduced. The efficiency of the normals in Table 8 compared with their efficiency in Table 4, sitting for sitting, amounts to the following: in the first sitting, 74.5% (i. e., they were 74.5% as efficient in Table 8 as in Table 4) ; in the second sitting, 76.2% ; in the third, 74%; in the fourth, 81',; ; and in the fifth, 82.8%. The corresponding figures for the epileptics — i. e., the efficiency in Table 9 compared with Table 5 — are as fol- lows: first sitting, 59.6%; second sitting, 69.6%; third sitting, 78% ; fourth sitting, 79' ; ; and fifth sitting 72.6' ! . It must be remembered that it was relatively easy to score perfect on one of the figures, since one of the figures contained three identical digits. But the efficiency in Tables 8 and 9 is very high relative to the efficiency in Tables 4 and 5, particularly in view of the fact that the scores from the perfectly reproduced figures are also contained in Tables 4 and 5. Relatively both the normals and the epileptics (with one 3b The Measurement of Mental Traits in exception each) gained from sitting to sitting in their ability to reproduce whole figures correctly. Relatively the normal children reproduced more whole figures than did the epileptics. That is to say, they scored relatively higher than the epileptics when the scoring was based only on the total credits than when it was based on tne combined total and partial credits. DEFERRED RECALL ( RETENTIVENESS) At the beginning of the March sitting the subjects were tested with respect to their retentiveness of the figures which they had tried to memorize one month earlier (A2). The figures, of course were not exposed again, nor had they been accessible during the month to any of the subjects. No in- timation w s given during the second sitting, or at any time, that such a test as this would be given. The time allowed for writing the figures was the same as in the original test. The results are given in Table 10. TABLE 10 Retention During March Test of Digits Memorized in February: A 2: Normal Children Total Number of Figures Completely Correct Epileptics Total Number of Figures Completely Correct No. Age Sex Average Number of Figures Completely Correct Subject Age Sex 2 7 Boy 11 6 .5 1 81 10 Boy 0. 0 2 7 Girl 6.6 .0 0 102 10 Girl 0. 0 4 7 Ave. 9.9 .3 104 10 Girl 0. 0 8 8 Boy 5.41 .37 3 99 11 Girl 0 0 3 8 Girl 8.3 5 1 79 12 Boy 6.6 0 11 8 Ave. 5 9 :4 84 12 Boy 3 9 Boy 9.96 .0 0 91 12 Boy 6.6 0 4 9 Girl 6.63 .33 1 108 12 Girl 0. 0 7 9 Ave. 9.6 .2 93 13 Boy 4 10 Boy 10.0 .5 Y 86 13 Boy- 0. "6' 3 10 Girl 11.06 .66 2 97 13 Girl 13.3 1 7 10 Ave. 10.64 .6 77 14 Boy 30.0 1 3 11 Boy 8.86 .66 2 89 14 Boy 0. 0 8 11 Girl .13.3 .5 4 80 15 Boy 0. 0 11 11 Ave. 12.0 .54 82 15 Boy 6.6 0 5 12 Boy 9.28 .C 0 >3 15 Boy 0. 0 3 12 Girl 13.3 .66 2 103 15 Girl 0. 0 8 12 Ave. 10.8 .25 78 16 Boy 0. 0 4 13 Boy 12.47 .5 2 88 16 Boy 2 13 Girl 14 9 .5 1 85 16 Boy 0. 0 6 13 Ave. 13 3 .5 96 16 Girl 0. 0 4 14 Boy 7.5 .5 2 98 16 Girl 0. 0 7 14 Girl 9.96 .2 1 106 16 Girl 0. 0 11 14 Ave. 8.8 .33 100 16 Girl 0. 0 2 15 Boy .0 .0 b" 87 17 Boy 0. 0 2 15 Girl 15.0 .5 i 90 17 Boy 0 0 4 15 Ave. 10.0 3 95 17 Girl 0. 0 1 16 Boy 105 18 Girl 0 0 3 16 Girl 15.5 ]33 i 107 21 Girl 6.6 0 4 16 Ave. 15.5 .33 101 24 Girl 0 0 1 17 Boy 6.6 .0 0 2 17 Girl 10 0 1.0 l 3 17 Ave. 8.3 .5 8.3 .33 n Average Boys 3.7 .07 11 7 .45 15 Avpratrp fiirls: 1.4 07 Average Boys and Girls ■ 10.0 .39 26 Ave. Boys and Girls. . 2.5 .07 Normal and Epileptic School Children 37 The average efficiency of the normals was 10%, which is only 24% of the efficiency which they showed in the test one month before. The aver- age efficiency of the epileptics was 2.5%, which is only 11% of the initial efficiency in the test. The epileptics showed only 25% of the normals' efficiency in the deferred reproduction as against 62%^ of the normals' efficiency in the immediate reproduction. In other words, the epileptics' capacity of retentiveness was, relatively, only half as good as their power of immediate recall, when measured by the normal capacity for immediate and deferred recall. The facts brought out by this test are well recognized by epileptologists and teachers of epileptics. Epileptics are exceedingly slow at "learning lessons" in school, and, what is worse, facts committed to memory and reproduced well at the time frequently cannot be retained for any length of time. We have found epileptics who made perfect scores in spelling im- mediately after the lesson had been studied, but who would fail on nearly all of the words after the lapse of a week. We have previously pointed out in connection with a study of the Binet scale,1 that "epileptics suffer from a fundamental impairment of memory, as shown by consistent failures in all the tests of memory span (sentence and number tests), by the inability to reproduce six units from reading a short passage once, by the inability to recall their ages, and to remember and execute three simple commissions." We found that the average number of units reproduced in the Binet reading selection by epileptics ranging from VII to XIII years mentally was only 5.2.2 Xot a single one of the \T-year-olds passed the Vl-year sentence memory test, only 27% of the VH-year-olds passed the 5-digit test, and only 2.2% of the IX-year-olds reproduced six memories from the reading test. The percentage of passes in most of the other tests in these ages was decidedly higher. We have, of course, tested epileptics who have had very good mem- ories, particularly in certain directions, but our experimental findings cor- roborate the observed fact that the average epileptic suffers from a very weak memory. The weakness is probably due both to the low state of impression- ability of the cortical neurones and to the destructive effects of repeated con- vulsions' which may not inappropriately be likened to ''brain storms" which sweep away the traces left by earlier impressions. It happens very frequently that the work with epileptics in the schools, particularly the literary work, must be done over and over again before any very permanent impressions are left. And then the results of our efforts are not infrequently entirely, or largely, obliterated by a severe seizure. We have known epileptics who, after a severe attack, had so completely forgotten how to use tools of which they had acquired considerable mastery, that the use of the tools had to be taught all over again. But our observations seem to show that abstract or literary attainments are more easily lost than motor skills. 'Experimental Studies of Mental Defectives, 1912, p. 53. 2P. 80. 38 The Measurement of Mental Traits in The normal girls did better in the deferred recall than the normal boys, the boys' efficiency being 71% of the girls'. In the immediate memory test, it will be recalled, the boys did a trifle better than the girls, the girls' efficiency being 98% of the boys'. Among the epileptics the boys did de- cidedly better in both the immediate and the deferred recall, but especially in the deferred recall, the girls' capacity in the deferred reproduction amounting to 38% of the boys' as against 80% in the immediate reproduction. In the sixth and eleventh columns are given the number of figures which were reproduced perfectly, that is. in which all three of the digits were recalled in the right order. Only two epileptics and 26 normals reproduced any figures which were completely correct. This gives an average of only .39 of a figure for the normals, and of .07 for the epileptics. Almost needless to say, the figure most frequently completely recalled was 999. Many gave 555, which had been shown in January on the first card, Al. Other numbers on Al were also occasionally mistakenly reproduced. Indeed, it is probable that the attempted reproduction of Al in February would have proved more successful than the reproduction of A2 in March, owing to the tendency to confuse the numbers which appeared on Al with the numbers which appeared on A2. 6. CONCLUSION'S The general conclusions throughout are offered tentatively, for reasons we have already given. 1. The capacity of epileptic children for memorizing digits is distinctly less than the capacity of normal children. The epileptics were inferior to the normals both in the immediate and the deferred recall of the digits, but especially in the deferred recall. While their power of immediate recall amounted to 62% of normal capacity their power of retentiveness after the lapse of one month amounted to only 25' , . 2. The normal boys were slightly superior to the normal girls in the immediate recall of memorized digits, but the girls distinctly excelled the boys in the deferred recall. In our dental squad the boys were slightly superior to the girls in the immediate recall. The epileptic boys were dis- tinctly superior to the epileptic girls in both the immediate and deferred recall. On the whole, the difference between our normal boys and girls is not very considerable, while our epileptic boys are distinctly superior to our epil- eptic girls. As we have previously intimated, it may be a matter of chance that our group of epileptic girls was inferior to our group of epileptic boys, althov~h there is some slight evidence that the females in institutions and speciai classes are inferior to the males.1 According to the school's rating, ■J. E. Wallace Wallin. Report of the Board of Education of the City of St. Louis, 1915, p. 148. Psycho-Motor Norms for Practical Diagnosis, 1916, Chapter IV. Normal and Epileptic School Children 39 our group of normal girls may have been slightly superior to our group of normal boys, for while 25.6' < of the girls were rated as dull compared with 24.3' ; of the boys, and while only 33.3% of the girls were rated as average compared with 43.2% for the boys, 41% of the girls were rated as bright compared with 32.4% of the boys. The difference in the intelligence status of the normals boys and girls, however, could not have been large. 3. The ability of the normal children in memorizing digits appears to increase slightly from age to age. The exceptions which we find may be accounted for by the fewness of the cases and the preponderance of dull and average pupils in the ages showing losses. The oldest pupils did 1.6 times as well as the youngest pupils. On the other hand, there is no consistent im- provement with increasing chronological age among our epileptics. This statement applies to all of the other tests. 4. The ability to recall digits memorized under the conditions of this test, is very considerably reduced after the lapse of one month, the normal children doing only one-fourth as well in the deferred recall as in the im- mediate recall. Had the cardboards not contained a figure with identical dig'ts the record in the deferred recall would have been still poorer. At the same time, the record in the deferred recall would have been better had an earlier series of numbers (A) not been shown. 5. The ability to recall figures with three identical digits is, of course, distinctly higher — at least 200' ; higher — among both normal and epileptic children, than the ability to recall figures with three different digits. The epileptics profit relatively somewhat more than the normals when the scoring is based exclusively on the figures with identical digits. In other words, the epileptics (as was to be expected) are relatively most inferior on the difficult combinations. The normal pupils also memorized the figures with the initial zeros somewhat better than they memorized all the figures, but not so the epileptics. 6. In absolute units the normal children gained more than the epil- eptics, but relatively the epileptics gained more than the normals. That is to say. the epileptics gained, more on their own initial or early records than the normals gained on their earlier records. This is in accordance with ex- pectation, for the normal pupils may be presumed to have initially func- tioned nearer their maximum than did the epileptics. But about one-third of the subjects made no gain or actually lost during the course of the experiment, the proportion being a trifle higher for the epileptics than for the normals. 40 The Measurement of Mental Traits in 7. The average monthly improvement during the five sittings was about the same for the normal boys as for the normal girls — in the earlier experi- ment with the dental squad the boys improved slightly more than the girls — while the epileptic girls improved more than the epileptic boys. 8. The amount of the improvement did not seem to vary with the chronological age of the children. 9. In order to use this type of an immediate memory test of non-sense materials (ten three-place digits visually presented during a 45-second ex- posure), it is necessary to base the scoring on each separate digit rather than merely on the three-place figure as a whole. When credit is given only for figures in which all three digits are correctly given, the unit of measure- ment is too large. Some pupils would receive zero for figures in which they would have only one or two digits correct. They would be deprived of partial credit for partially correct answers-. As a matter of fact, if we dis- regard the figures with identical digits many pupils were unable to make a perfect score on any figure in one or more of the sittings. We have seen, however, that the efficiency when based on total or perfect credits only (Tables 8 and 9) amounted in most sittings to over 75% of the efficiency when based on the combined sums of partial and total scores (Tables 4 and 5). That this test is by no means an easy one, is indicated further by the fact that the average score for the normal children was only about 41% of the possible score. In our previous use of the test with 27 normal children, ages 10 to 16, who were receiving dental treatment, the average score for the entire series (six) was 2.7% higher, a negligible difference. The dental group did better in two ages that could be compared and our normal group in five ages. In the dental experiment, however, the sittings were distributed at longer intervals throughout a whole year.1 The highest score made in any one of these sets of figures by any one whom we have tested is 90%. The highest score made in this experiment was 83.3%. The highest number of figures reproduced completely correct (all three digits correct) by any one whom we have tested is eight — by the person who scored 90% . The above figures refer to the immediate recall. The individual differences in the test are very considerable. The high- est and the lowest individual scores made by the epileptics in the immediate recall are as follows: in sitting 1, 46.6% and 0; in 2, 43.3% and 0; in 3, 46.6% and 0; in 4, 53.3% and 10% ; and in 5, 56.6%0 and 0. The individual extremes among the normals in the corresponding sittings are: 80% and 6.6%; 80%) and 10%; 73.3% and 16.6%; 83.3% and 16.6% ; and 80% and 10%. It should be remembered that the normal pupils differed in age from 7 to 17 years. The individual differences in each age (the difference between 'Dental Cosmos, 1912, p. 404f. and 545f. Normal and Epileptic School Children 41 the lowest and the highest scores made by the normal pupils of a given age), are as follows, based on the average figures for the five sittings (1-5): Age 7 8 9 10 11 12 13 14 15 16 17 Difference... 31.3 31.7 26.7 37.3 26.0 23.2 26.7 46.3 25.9 13.0 20.0 We asked two epileptic sections and one of the normal sections to express themselves on the question as to which one of the different tests given in this experiment was the most difficult one and which the easiest. Some of the subjects did not vote, because they were unable to decide. The epileptic boys were asked the question during two successive sittings, but the votes did not entirely correspond. Crossing off the A's (test E) was said to be the easiest by twelve epileptic boys, seven epileptic girls and one normal child. Reproducing the sequents in test G was declared easiest by one epileptic boy, constructing a sentence containing three supplied words (test J), and adding one-place digits (test C) by two epileptic boys each, and memorizing the figures in test A by four epileptic boys. The spontaneous association test ( B ) was thought to be the easiest by four normal pupils and the opposites test (D) by eleven normal pupils. Among the hardest tests were mentioned the addition test (C), by five epileptic boys, the range of observation test (F), by two epileptic boys and 11 normal children, the construction of sentences (J), by three epileptic girls, the memorizing of the figures (test A) by six epileptic girls, and the ink blot test (H) by three epileptic girls. This partial census did not indicate clearly which test was introspectively the hardest or which was the easiest. 10. In spite of the difficulties which beset the test, the improvement in the performance from age to age indicates that the test can be profitably used as one of a number of group tests for measuring intelligence or mental maturity. Used as a clinical test in individual diagnosis it would probably be better to allow the subject to use verbal instead of written responses. If so used norms should be established by clinical tests. III. RAPIDITY OF THINKING Spontaneous Association with Supplied Key Words: Bl to B5 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS The stimulus material consisted of the printed blanks. 4 x 14 inches, each containing 30 words in 12 point type. The stimulus sheets also served as record sheets. The first three blanks are reproduced: Book House Curtain Piano Horse Hay Leather Slate Buggy Flag Rain Winter Apple Cleveland Nose Canary Bread Chalk Carpet Mountain Recess Father Roosevelt Addition Coal Lying Ball-game Sunrise Church Lungs Bl Table Bed Window Music Dog Grass Shoes Top Wagon Star Hail Summer Orange Chicago Mouth Robin Butter Blackboard Ruj River Reading Mother TaftT Substraction Gas Stealing Work Noon School Stomach B2 42 Pencil Chair Glass Singing Cat Potatoes Knife Skates Automobile Moon Storm Spring Peach New York Head Rooster Toast Dictionary Ceiling Vallev Spelling Brother Harmon Division Gold Laziness Study Evening Factory Heart B3 Normal and Epileptic School Children 43 The 180 words used in the complete set (for reasons already given, only five of the sets could be used in this experiment) were selected from a very much larger preliminary list of familiar words. From this preliminary list 180 words were selected and arranged by the writer in six lists of 30 words each. The attempt was made to make each list approximately equal in total difficulty, by distributing an equal number of easy and hard words on each sheet. Practically all of the words selected were common nouns. Each sheet contains only one verb and two proper nouns, one being the name of a well-known city (except on the sixth sheet) and one the name of a well- known person (an assumption, no doubt, justified so far as concerns the group of Cleveland children who were first tested). These six lists were then submitted to three instructors in a training school for teachers, who were asked to rearrange them into six lists, each of which should represent col- lectively the same degree of difficulty. The final lists were based on the combined judgments of all who thus assorted the words. This method of constructing word lists of supposed equal difficulty is, of course, open to various objections. None of those who arranged the lists felt a very high degree of confidence in the correctness of his arrangement. But it is probable that the combined judgments of several persons is superior to the judgment of one person. To obtain thoroughly satisfactory lists we should probably have to test six groups of children of equal degrees of capacity, as stated in the preceding section — it was impossible to do this at the time these lists were prepared — but. even so. the difficulty of the different lists would probably vary more or less according to the varying experiences of different groups of children. The readiness of words of the same degree of familiarity to arouse associations will inevitably differ among different individuals according to their varying experiences, a condition that can scarcely be remedied. But we would emphasize that these word lists, even though they should prove not to be equally difficult, can be legitimately used for the purpose for which we are especially employing them, namely for gauging the comparative efficiency of our normal and epileptic subjects. Our tests are somewhat marred by the inclusion of a few local words. The subjects, however, were always instructed to pass over words whose meaning they did not comprehend or words which did not suggest any words to them. 2. NATURE AND PURPOSE OF THE TEST This is a species of '"free" or spontaneous association test. It is free in the sense that the subject was permitted to write any words at all, logical or illogical (as a matter of fact, the vast majority of the words written were log- ically connected with the antecedent key-words), suggested by each of the 44 The Measurement of Mental Traits in supplied key-words. The test differs from the ordinary "running" free association test, in which the subject starts with only one supplied key-word and writes or recites a series of words, each one of which is suggested by the word which has just been uttered or written. Either one of these so-called association tests may be used as a test of the speed of ideating or thinking, since thinking is essentially a process of forming connections or associations between ideas, and of expressing such associations in words. There is prob- ably little thinking that is entirely imageless or wordless. The number of words written in this test represents the rate of forming spontaneous associations with predetermined antecedents plus the time re- quired to write the words. The writing time is, of course, a considerable factor. Most subjects can utter word associates very much faster than they can write them. Moreover, the writing time complicates the experimental problem, because children differ a great deal in the speed of writing who might not differ very greatly in the speed of forming associations. The younger children write much slower than the older children, while many epi- leptics whom we have 'studied and probably many backward children write much slower than normal children. The uncertainty or inability to spell the word, would serve to further retard the responses of the younger, backward and epileptic children. In order to mitigate this retarding influence the children were instructed not to linger over the spelling of words — not only in this test, but in all the tests — in order not to lose any time. Some children wrote extremely few words. It would be difficult to say without investigating each case whether this was due to inability to spell, difficulty to write, or slowness in ideating the words. A few children occasionally wrote two or three words, or a phrase of two or three words. Such responses were given only one unit of credit. From what we have said, it is probable that had the words which sug- gested themselves to the subjects been expressed verbally instead of in writ- ing the younger children and the epileptics, and possibly the backward chil- dren, would have done relatively better than they did, although it is likely that most younger, backward artd epileptic children also ideate slower than normal children. It is further evident that if some children tended to write long words and other children short words, the latter children would probably be able to write more words. Moreover, children who made the greatest progress in writing during the five months of the experiment, or who tended progressively to write shorter words would tend to show the greatest improvement in the test. If the test is to be used to measure the amount of gain made in form- ing associations we must assume that the writing time remains approximately constant from test to test, or that the average length of the words was the Normal and Epileptic School Children 45 same in the different tests. The impression gained from the examination of the successive lists seemed, in the main, to justify the latter assumption. It must be remembered, therefore, that the time in this test represents both the time to think and the time to write the suggested words, since it is impossible under the conditions of the experiment to measure independently these two variables. The necessity of giving group tests limited the reactions to written responses. In our discussion of the results the above considerations must be borne in mind. 3. DIRECTION'S FOR GIVING THE TEST General Instructions: Follow the general instructions on page 15. Order of Giving Tests: During the first sitting, Bl, Second sitting, B2, etc. Time: 60 seconds (one minute). Pupils sit with folded hands. Have the papers placed upside down on the desks occupied by the pupils, with the request that they must not touch them. Instructions to be Given to the Pupils: "On the other side of the paper you will find a list of simple words. When I say 'turn' you must turn your papers at once and write opposite each word the first word that it makes you think of. You must try to write the first word that comes to your mind, whether it makes any sense or not. Don't write the same word and don't write sentences; just single words. If you come to a hard word that doesn't make you think of anything, don't stop, but go right on to the next word. Later on if there is time you can go back to this word. If you do not know how to spell a word that you want to write, spell it as well as you can so that you will not lose any time. Nothing will be counted off for words spelled wrongly. (Illustrate test on board by means of the following words. Use no others in this or later tests: Book — paper. House — wood. Curtain — window. Piano — music, or singing. Horse — drive, or animal.) "Now. ready!" "Turn your papers!" " Now, stop!" ' Turn your papers!" Signature: At the close of the test the pupils write at the top of the paper: full name, age, sex, date, hour, condition, school, and grade. Grading of Papers: The ranking will be on the basis of the number of words (associates) written. Each word will be rated 3.3%. Ten per cent should be given for three associations. Find the per cent of efficiency (in decimals, not fractions). The efficiency is based on the speed and not on the quality of the associations. 46 The Measurement of Mental Traits in 4. TABLES TABLE 11 Spontaneous Association with Supplied Key-words: Bl— 5 Normal Children INO. Ae. Sx 1 % S Vc f\ 4 f 4 ft f 5 / 4 2-8 % 4- 9 m f\ 2- 9 X Pain Ave V 2 7 B 21 6 15 u 0 1 LI c 0 1 . * AA on n u 92 on £AJ n u 1 00 1 Q . c ■5 O 20 0 i Us 1 OO 19 1 1 6 2 7 G 41.6 33 3 O O 1 1 Dv OG 9 1 13 2y 9 1 00 OA £if 1 1 29 9 i i 1 u 1 1 CiH V\J 30 3 2 0 4 7 Ae 316 21 1 27 4 1 1 A*? iy 1.06 19 9 1 00 O ') 1^1 7 19 9 i i in 1 1 . 11 24 7 IS g g B 25.2 26 6 27 6 1 Oil 33 2 1 07 29 1 89 (Jo Q O 30 1 i .uz 1 57 0/ 29 0 '9 3 8 G 13 3 17 no £.£. 1 1 oz £o n 9 1 54 oh A V .67 1 u Q 24 9 i i 1 7 93 1 i CO 22 4 .7 8 Ae 22.0 24 2 25 9 1 73 32 2 1 21 26 6 83 26 28 7 i 13 49 27 2 8 3 g B 14 4 25 0 33 3 99 32 2 1 06 33 3 1 00 33 f 32 7 i 01 i 02 32 8 4 4 g G 30 8 23 3 29 9 95 23 8 94 2 1 8 1 11 24 5 23 2 90 88 23 9 — 1 8 7 9 Ae. 23.7 23 9 31 6 97 28 5 99 26 1 1 06 28 5 27 3 95 93 27 7 — 8 4 10 B 39 \\ S8 8 38 8 1 03 45 8 1 18 45 7 1 00 38 7 45 7 i 18 i 19 43 0 2 9 3 10 G 42 2 39 9 43 3 1 22 37 7 87 38 8 95 41 6 38 3 90 95 39 9 — 3 7 10 Ae. 40.4 39 4 41 0 1 15 42 3 1.05 42 8 .98 39 9 42 5 i 06 i 08 41 7 1.4 3 11 B 39.9 36 6 46 6 1 32 39 9 .86 41 0 1 02 41 6 40 5 98 i 16 41 0 1.4 8 11 G 47.8 39 9 50 3 1 29 47 8 .93 46 6 1 .03 45 1 48 0 i 04 i 34 49 1 2.4 11 11 Ae. 45.7 39 0 49 3 1 30 45 7 ■ 91 44 9 1 .03 44 1 AR A u i 03 i 29 AC 46 9 2 1 5 12 B 51.9 46 6 55 9 1 19 59 9 98 57 3 1 03 51 2 56 6 i 09 i 22 53 9 3 8 3 12 G 51 0 50 1! 57 7 1 16 58 8 1 AO 52 2 QQ OO 53 8 55 C 0 i 02 i 03 ~A 04 0 8 12 Ae. 51.6 47 9 56 6 1 18 59 5 1.00 55 4 .97 52 2 56 2 i 07 i 15 54 l 2.6 4 13 B 52 2 44 4 51 6 1 16 50 8 .98 55 8 1.15 48 7 53 2 i 08 i 20 50 6 3 3 2 13 G 41 6 38 3 48 3 1 27 41 6 .79 33 3 1.17 43 3 37 4 84 87 40 3 - 1.6 6 13 Ae. 48 7 41 9 50 5 1 21 47 7 .91 48 3 1 15 46 9 48 0 i 00 i 09 47 2 16 4 14 B 44 2 54 1 58 3 1 07 62 2 1.09 57 7 94 56 2 59 9 i 07 i 07 57 4 1.6 7 14 G 42 8 54 7 54 6 98 52 3 .97 56 1 1.08 53 5 54 2 i 00 i 04 54 2 5 11 14 Ae. 43.3 54 5 56 2 1 02 55 2 1.01 56 6 1.04 54 6 56 3 i 03 i 05 55 4 8 2 15 B 51.3 68 3 93 3 1 12 75 1 04 66 H .90 70 8 70 8 i 02 99 71 3 6 2 15 G 50. 1 63 3 68 3 1 08 74 9 1.09 .69 9 96 65 8 72 4 i 09 i 10 79 1 2.2 4 15 Ae. 50 7 65 8 76 6 1 09 74 9 1.07 68 3 .93 68 3 71 6 i 05 i 05 75 2 1.5 1 16 B 53.4 56 6 50 0 88 53 3 1 06 56 6 51 6 91 91 53 3 - 16 3 16 G 56.8 61 6 61 b i 13 72 1 1.22 66 6 .92 58 3 89 4 i 22 i 23 65 1 1 2 4 16 Ae 55.9 59 9 til 0 i 13 66 6 1 13 63 3 96 57 9 64 9 i U i 15 62 2 6 1 17 B 30.6 50 9 50 0 i 00 50 0 1.00 66 6 1.33 50 0 58 3 i 16 i 33 54 1 5.5 2 17 G 44 4 66 6 68 3 i 00 65 0 .95 66 6 1.03 68 3 65 8 96 97 7 2.0 3 17 Ae. 39.8 58 3 62 2 l 00 60 0 .96 66 6 1 13 62 2 63 3 i 03 i 09 62 5 3.3 Ave . B. 37.7 39 5 41 i i 34 45 6 1.03 44 6 1.00 41 7 45 2 i 06 26 43 5 2.0 Ave . G.. 42.3 43.4 48 7 l 17 47 9 1.02 45 6 .99 46 1 46 8 i 04 1 10 46 5 8 Ave . B&G . 40.1 41.4 46 .5 l 25 46 9 1 02 45 fi .99 44 0 46 0 i 05 17 45 1 1 4 Normal and Epileptic School Children 47 TABLE 0 Spontaneous Association with Supplied Key-\Vords» Bl— 5 Epileptics Su. 1 Ae. Sx. 1 ,1 3 4 % fi 5 fi V3 4~ 5 % _ 2 ~5 Gain Ave. % 81 10 ■ B 0 * ■ • • • • ■ 6.6 — r- 13 3 2 01 0 0 y .y ft Q y .y _ b. 7 102 10 G 6.6 20 0 6.6 99 OO 16.6 9 C1 c Oi 0-0 4U 19 9 10 0 1 1 fi 110 ■Of 99 , OO 19 A ic .4 — 4 4 104 10 G 27 6 40.0 30.0 .75 60.0 C . UU 56-6 94 9C n 00 U Cfl 9 00 .0 1 fifi 1 . 00 1 A 1 1.41 AC fi 40 0 5-5 • " • Av. G 17 1 30.0 18 3 .54 38 3 9 9C C .CO 31 6 fi7 9J 1 Z4 1 9J Q 04 y 1 9fi 1 .iO ft7 0/ 9Q C cV 0 c . O 99 11 G 13 8 43 3 33.3 . 76 — JO 0 i in 1 1U 9Q 9 00 .0 9fi fi 00 .0 Q9 .yo HA .04 97 7 0/ / 9 1 — O .0 79 12 B 9.0 3.3 23 3 7 06 6.6 9fi .CO 6 6 1 .00 19 9 10 .0 fi £ 0.0 AO. 4y 9 ftn c .UU 9.9 1 1 84 12 H 16 2 26.6 23 3 87 26.6 1 \A 1 . 14 26.6 1 00 OA O Z4.y 9fi fi iO.O 1 fifi 1 .uo 1 ftft 1 UU 9C Si CD O 0 . 91 12 | 18.6 23.3 23 3 1 .00 33.3 1 49 1 AC 23.3 .69 23 3 9D 9 1 91 I .cl 1 ftft 1 UU CO . 0 0. Av . 5 14.6 17.7 23 3 2 .97 22 1 QA 18.8 .oy 90 C C\J .0 OCl C C\J . 0 Q9 . y£ 1 99 1 .00 9ft fi iU .0 1 . 0 108 12 § 0 0. 0. ••• • 0. 6 6 • ■ * ■ U. 9 9 O . O 9 9ft O . OU 1 fi 1 .0 9 9 c c 93 13 B 3.3 6 6 23.3 3 53 20 0 QC .00 20.0 1 00 is n I. U 9n n U 1 99 1 OO 9 ft9 O .Uo 17 .5 4 4 QC 13 B 9.0 13 3 16 6 1 .24 13.3 QA 6.6 ao. .4y 10 u 1ft ft 1U u fifi .00 AO. 4y 1 9 A ic 4 0 9 — c . c A *■ Av. B 6.1 9.9 19 9 2 38 16 6 fl9 13.3 .74 10 . u 1 c ft 10 u QQ yy 1 ~fi 1 . /0 14 .9 1 . 1 _ 97 13 G 6.6 13 3 16.6 1 24 6.6 .4U 10 u fi fi 0.0 A A .44 4y 19 9 ic C Q 9 — O O 77 14 B 23 4 36 6 33 3 90 33 3 i ftft 1 .UU 40 0 1 .20 00 . u 9fi fi 00. 0 1 ft* 1 .U4 1 ftQ 1 uy Or q OO O 1 1 89 14 B 0. 20 0 33 3 1 66 30.0 on .yu 23.3 .77 OC fi CO 0 9C C zo 0 1 .00 1 1 fi 1 10 26 6 1 1 Av. B 11 7 28 3 33.3 1 .28 31 6 qc yo 3l 6 yo 9fi 9 CO 6 91 fi ol 0 1 19 1 . lo 1 19 1 IC 31.2 1 . 1 80 15 B 0. 16 6 10.0 .60 20 0 2 00 10.0 50 13 3 15 .0 1 19 1 Ic fift 14 . 1 — 2 2 82 15 B 11.4 20.0 26 6 1 33 20 0 7C to 26 6 1 .33 99 9 CO O 99 9 CO O 1 ftft 1 UU 1 99 1 .00 99 9 co .0 2.2 83 15 B 0 10 0 10 0 1 .00 6.6 DO 16 6 2 51 ift ft 1U U 11 fi 11.0 1 lfi 1 10 1 fifi 1 00 10 8 2.2 • ■ • ■ Av. B 3.8 15 5 15.5 97 15 5 1 . 10 1.44 ice 10 • 0 lfi fi 10.0 1 ftQ 1 uy 1 1U 1 iy 16 .0 ■ 7 103 15 G 0 0. 6.6 ■ * * ' 10 0 1 C 1 1 01 13.3 1 33 3 .3 11 fi 1 1 0 1 ^1 0 .01 7.5 4.4 78 16 B 0. 6 6 13 3 2.01 16.6 ] .24 10.0 .60 10 0 13.3 1 33 1 CI 1 51 116 1 1 88 16 B 16 6 23.3 13.3 57 20 0 1.50 23.3 1.16 18 3 21.6 1 18 1.00 20 0 0. 85 16 B 9 0 16 6 30.0 1.80 20 0 .66 33.3 1 r6 23.3 26 6 1.14 2.00 25 0 5.5 • • ■ ■ Av. B 8.4 15 5 18.8 1 .46 18.8 1 . 13 22.2 1 14 17 2 Oft z CV.O 1 .21 1 .50 18.8 2.2 96 16 G 0. 0. 0 6 6 0. fi c 0 b 2 2 9 1 0.0 98 16 G 9 9 13.3 13 3 i 66 6.6 !49 13 0 10 0 ''75 11 1 -3 3 106 16 G 6.6 13.3 16.6 1.24 10 0 .60 6.6 .66 15 0 8.3 .55 49 11 6 —2 2 100 16 G 0 10 0 13.3 1.33 0. .li 13.0 11.6 6.5 56 1 30 9 3 1 Av G 4.1 7.7 10.8 1 28 7 7 53 8.2 '57 9.5 7.8 .87 89 81 - .3 87 ! 1" B 0. 10 0 0. .0 0 0 5- 0 0 0. 2.5 -3.3 90 ! 17 B 0. 0. 30.0 0 '6' 0- 15. 0. .0 0. 7.5 0 Av B 0 5. 15. ;;;; 0. • 0- 10. 0. 0 0 5.0 —1.6 95 1 17 G 0 0. 0 0 0. 0. 0 0 0 .0 0 105 ' 18 G 6.6 0. 10. 0. 13.3 5. 6.6 1.32 5.8 4.4 107 21 G 9 0 10 10. 1.00 3.3 ! .33 6 6 2 00 10. 5. 50 ^66 7 5 -11 101 1 24 G 0 10 6.6 .66 0 | 0 3 3 8 3 1.6 .19 .33 5 0 -2 2 Ave B 7.2 15 5 20.6 1.68 17 0 j .95 17.4 1 13 18.3 ; 17 2 91 1 19 17 6 9 Ave . G 6 1 12 3 11.6 l .91 10.3 88 13.0 91 11 9 : 11.8 1.12 .65 11.8 2 Ave B & G 6 7 14 0 16 2 , 1 40 14 3 1 92 15 3 1.05 15-5 13.1 1.01 .98 15 0 .6 5. RESULTS (See Tables 11 and 12) The evaluation of the results in this test might be made from two points of view, first, from the standpoint of the quality of the associations, and; second, from the standpoint of the number of associations which were formed. Our discussion is limited entirely to the quantitative aspect, owing to the d"ff.culty of making a satisfactory graded qualitative evaluation of the re- sponses. We repeat, however, that the vast majority of the associations were of the logical type. Biriet and Simon conclude from a different associa- tion test "that logic is easier to follow than chance.'' COMPARATIVE EFFICIENCY The average efficiency, based on the mean for all the subjects and for all the sittings except the first1 (column headed %*) was 45.1% for the normals 'The scores made in the first sitting are not included as the time allowed during this sitting (85 sec.) was inadvertently lengthened 25 sec. 48 The Measurement of Mental Traits in and 15.0' t 'for the epileptics. In other words, the epileptics' efficiency amounted to only 33% of the normals. The epileptics did only a little more than one-half as well as the youngest normals tested, and not quite one- fourth as well as the oldest normals. We have previously called attention to the pronounced psychic and psycho-motor retardation affecting the reactions of epileptics.1 This retardation appears more emphatically in the present group test than in any of the previous clinical tests which we have given, possibly due, in part at least, to the fact that the average normal child is able to write more rapidly than the epileptic child. We shall see later from test L that the speed of executing a very simple writing movement, writing circles, was slower for the epileptics than for the normals. The boys' capacity, based on the general average at the bottom of the column for all four sittings, was slightly lower than the girls' among the normals, but considerably higher among the epileptics. The normal boys did 93' "X as well as the normal girls, while the epileptic girls did only 67% as well as the epileptic boys. The girls did better in every sitting among the normals and poorer in every sitting among the epileptics, based on the sex averages at the bottom of the percentage columns. Based on the sex averages for sittings 2 to 5, the normal girls did better than the normal boys in six ages and poorer in five ages, while the epileptic girls did better than the epileptic boys in only one age and poorer in five ages, no comparative data being available in the other ages contained in the table. Thus, while the difference between our normal boys and girls is not very significant, the epileptic boys are distinctly superior to the epileptic girls. This may be due. as we have already intimated to the fact that we were dealing with a group of epileptic girls who were more than ordinarily deficient. Considering the fewness of the subjects there is a very consistent improve- ment in this test from age to age among the normals. Based on the averages of the two sexes for sittings 2 to 5. there is an improvement in eight and a loss in only two ascending ages, when the score in each age is compared with the score in the next higher age. The loss occurs in age 13, in which half of the children were classed as bright, half as average, and nene as dull, while in age 12, 62% were classed as dull and average;, and in age 16, in which three-fourths were classed as dull or average. The score in age 17, in which all the pupils were classed as dull or average, is practically the same as in age 16. The highest score is in age 15, in which all of the pupils were classed as bright or average, and three-fourths as bright. The difference between the averages in the lowest and in the highest age amounts to 37.8% (62.5% — 24.7%). We have also found a consistent increase in the related Binet- Simon 60-word association test when epileptics are classified according to 'Experimental Studies of Mental Defectives, 1912, p. 106. Psycho-Motor Norms for Practical Diagnosis, 1916, p. 37. Normal and Epileptic School Children 49 Binet-Simon age.1 The figures for the epileptics in the present test in which the data are arranged according to chronological age, indicate that there is no consistent improvement from age to age. In fact, the records of the oldest children are considerably inferior to the records of the youngest children. One epileptic (No. 95) failed to score at all in any sitting, three failed completely in four sittings, one in three sittings, four in two sittings, and five in one sitting. Only two or three of these failures could be directly ascribed to an earlier convulsion. One normal child failed in all the sittings (No. 5, a dull, restless, inattentive boy of seven), and two in the first sitting, an •average" boy (No. 7) of eight, and a "dull" boy (No. 43) of nine. The normals made the highest score in the fourth sitting and the epi- leptics in the third, while both the normals and the epileptics made the lowest score in the first sitting. COMPARATIVE IMPROVEMENT The average monthly gain made by each subject from sittings 2 to 5 (last column in the tables) amounted, in absolute units, to only 1.4% for the normals and .6% for the epileptics. The epileptics improved 43% as much as the normals. Since, however, the above gains are based upon absolute units, it is to be expected that the epileptics would do relatively better when the improvement is computed as a fraction of the size of the scores which they made, because the epileptic scores were considerably smaller than those made by the normal pupils. Relatively to the size of the scores the epileptics do better, but. on the whole, do not quite equal the normals, as may be shown by comparing the various indices of improvement, which are obtained by dividing the early scores into the later scores. The gain in the fifth sitting over the second sitting (/:}) amounted to 17% for the normals, while the epileptics suffered a loss of 2' ', , a difference of 19% in favor of the normals. The improvement, when based on a comparison of the average of the last two sittings with the average of the second and third (/££), amounts to 5% for the normals but only 1% for the epileptics, a difference of 4% in favor of the normals. The following are the percentages of improvement made in each of the three later sittings when the record in a given sitting is compared with the record in the next preceding sitting: Index $ £ | Normals '. 25% 2% -1% ' Epileptics 40% [\ -8% 5% The difference in favor of the epileptics amounts to 15% in the third sitting and 6'v in the fifth sitting, while the difference in favor of the normal h — t-.-i • ! ; : ' i .....'!■• '.. -i ; 'Experimental Studies of Mental Defectives, 1912, p. 78. M T 4 50 The Measurement of Mental Traits in pupils amounts to 10% in the fourth sitting. The greatest gain was made in the third sitting by both groups. Both groups suffered loss in one sitting the normals in the fifth and the epileptics in the fourth. The variation from sitting to sitting is greater for the epileptics. Many epileptics varied ex- tremely from sitting to sitting both in this test and in the other tests, as may be seen by an examinaion of the tables. Based on the average monthly absolute gain for sittings 2 to 5, nine epileptics lost and two made no improvement, which is 36% of all the epi lept.cs, while the corresponding figures for the normal children are nineteen and six, which is 33% of the normals. The absolute average monthly gain was about two and a half times greater for the normal boys than for the normal girls (2.% vs. .8%) and four and a half times greater for the epileptic boys than for the epileptic girls ( .» 7c vs. .2%). Based on the average indices of improvement at the bottom of the tables, the normal boys gained more than the normal girls in all of the indices, while the epileptic boys gained more in three indices, and lost less in one than the epileptic girls, and the epileptic girls gained more in one index Ihe gain in the fifth sitting compared with the second (/ f ) was 16' , greater for the normal boys than for the normal girls (26% vs. 10%), and 54% greater for the epileptic boys than for the girls (19% vs. a loss of 35%) The gam based on the average of the last two sittings compared with the average of the second and third 4-5 % n-i * .5 f\ i - Gain Ave I-S 2 7 B 0 2.0 6 3 3 00 4 0 .66 .0 .00 1 0 2.0 2.00 2 4 0 2 7 G 8 I) 4.0 33 4 0 1 00 8 0 1 00 2.0 .25 7 0 5 70 !l6 5 3 -1.7 4 7 Av. 4 0 2.6 5 0 1 66 ,6 0 .83 1.0 .16 4 0 3.5 1 13 3 8 - .8 8 8 B 5 7 4 7 76 5 7 1 81 5 1 1.10 5.0 .80 5 2 5.0 1.26 i!6i 5 2 - .2 3 8 G 5 3 10.6 1 76 8 0 1 ( 2 9 3 1 27 5.3 .33 8 0 7.3 .88 1.00 7 7 0 11 8 Av. 5 6 6 3 88 6 4 1 48 6 4 1.16 5.0 .62 6 0 5.6 1 14 1 01 5 8 - .1 3 9 B 6 6 2.0 1 00 7 3 75 4 6 1.41 7.3 1.77 6 6 6. .81 .75 6 3 1 4 9 G 3 2.6 1 00 4 0 1 00 3 0 .25 6.0 1.75 3 5 4.5 .88 1.66 3 9 8 7 9 Av. 4 5 2.4 1 oo 5 6 90 3 7 .95 6.5 1 76 4 8 5.1 .85 1.30 4 9 .5 4 10 B 15 0 18.6 1 22 16 0 88 16 0 1.25 15.0 .9.5 15 5 15.5 .98 1.03 15 6 0 3 10 G 10 6 6.6 6b 5 3 75 6 6 1.83 9.3 1.50 8 6 8.0 .83 .94 7 7 - .3 7 10 Av. 13 1 12.6 94 10 6 83 12 0 1 44 12. S 1.13 12 5 12.0 .91 .99 12 2 • 1 3 11 B 20 0 14.6 81 9 3 78 12 0 1.88 9.3 .91 17 3 10.6 65 50 13 0 0 8 11 G 14 5 13.5 1 03 12 0 83 14 5 1.20 13.1 .86 14 0 14.0 1.09 1 26 13 fi .4 11 11 Av. 15 0 13 8 96 11 2 81 13 8 1.41 12.0 .88 14 9 13.0 .96 1 01 13 4 2 5 12 B 18 4 15 2 98 17 6 1 11 18 0 67 8 0 .32 16 8 12 0 72 64 15 1 -4. 3 12 G 17 3 14.6 1 19 21 3 1 48 16 0 .86 24.0 1.48 16 0 20.0 1.59 2 17 18 6 1.6 Q 0 12 AV. 18 0 AO .u 1 wo 19 0 1 25 17 1 . to 14 0 76 16 5 15 0 1 05 1 21 16 4 1 8 4 13 B 20 0 25.3 1 20 27 0 1 21 26 0 .85 25^0 L02 23 0 25^5 1.00 112 24 8 1 3 2 13 G 8 0 8.0 1 no 8 0 1 00 8 0 1.00 10.0 1.25 8 0 9 0 56 1.25 8 4 .5 6 13 Av. 16 0 18.4 1 15 20 6 1 17 20 0 88 20.0 1.06 18 0 20.0 .85 1 15 19 3 1.0 4 14 B 19 0 15.0 64 22 0 1 49 32 0 1.22 10 6 .61 17 0 21.5 1.09 .97 19 6 0 7 14 G 25 7 26 0 92 28 8 1 39 30 8 1 65 24 0 1 18 9.5 1 27 4 1 11 78 26 6 — .4 11 14 Av. 23 2 2L0 76 25 7 1 43 31 2 1.54 20^0 1.06 22 1 25 3 110 .84 24 0 • .3 2 15 B 31' 0 26.0 83 44 0 1 83 42 0 1.18 32 0 .76 29 0 37.0 1.27 1.00 33 8 0. 2 15 G 22 0 16.0 67 28 0 2 00 28 0 1 04 30.0 1 04 19 0 29.0 1 60 1.33 24 8 2. 4 15 Av. 27 0 21 0 75 33 n 1 94 35 0 1.11 31 0 .90 24 0 33 0 1.43 1.16 29 3 1.0 1 16 B 8 0 12.0 1 50 20 0 1.66 20.0 1.00 10 0 20.0 2.00 2.50 15 0 4 3 16 G 29 3 22.0 63 12 0 50 13 3 1.35 17.3 1.20 22 0 15.3 .62 .52 17 8 -3.5 4 16 Av. 24 0 18 6 92 12 0 50 15 0 1.45 18 0 1.15 19 0 16 5 97 1 01 17 1 -1 7 1 17 B 36 0 36.0 1 00 52 0 1 44 14 0 .84 44.0 1.00 36 0 44.0 1 22 1 22 42 4 2 2 17 G 16 0 4.0 1 00 14 0 92 20 0 2 16 14 0 62 15 0 17.0 1.27 92 15 9 - 5 3 17 Av. 22 6 20.0 1 00 26 6 1 09 28 0 1.72 24.0 .75 22 0 26.0 1 25 1.02 24 7 3 Ave. B 14 6 13.4 93 16 0 1 20 17 c 1.19 12.5 .90 14 .0 14.6 1.05 .95 14 7 — 3 Ave. G. 15 4 13.8 so 13 9 1 07 15 5 1 30 14 8 1 03 14 15.2 1.03 1.09 14 7 - 1 Ave. B & G 15 0 13*. 6 92 14 9 1 13 16 2 1.24 13 7 .97 14 A 14.9 1 04 1.02 14 7 - .2 Normal and Epileptic School Children 59 TABLE 16 Addition of One-Place Digits: CI— 5: Scores Based on Columns Correctly Added Epileptics Su. Ae. Sx. l % J h / j % i 3 % /I /c ft 5 70 f\ 1-2 4"5 ; 4-5 1-t / 1 1 ~5 % Gain Ave. 121 . R A U o .00 0 2 1 3 0 1ft9 in Q 4 A ** 00 V .00 g 2 [oo 4 6 1 !50 2'00 4 0 1 in \J A 4 00 1 00 o .00 4 4 2 50 1 00 3 2 0 Av nv. c 4 4 \ 00 £ 50 2 g 2.00 4 4 1 00 1 50 3 6 5 1 [ Q n u o - ft o 0 0 0 0. 70 g ft Q 4 0 2 1 0 1.3 O* O** 1 9 g ft Q 0 g 2.00 20 2.50 20 1.00 2 20 10 00 10 4 5 01 1° 1*. g o 1 25 16 g 00 20 1 .25 fcO 5 18 3 6u 2.00 12 4 2 Av. R 9 fi o ■ 25 g 5 00 14 6 1 87 18 .90 2.3 16 6 80 8.4 2.9 ins IUO 19 ft U U n o* n 0 0 !oo 0 0 93 13 g ft u ft o o o 0 0 !oo 0 0. 00 13 g ft ft 4 g 2 |o6 o .00 0 4 2.4 2.4 0. A u rv v . g 2 4 2 00 2 1 'i 0- 07 M xo J. n u 00 o 12 2 g 3 00 3 00 4 0 2 6 77 14 g 1 g 2 00 50 4 1.00 4 1.00 g 4 1 00 4 8 0. 89 14 g q 00 4 g 2 loO o !oo 2 4 2 00 00 3 2 -1. Av H 1 00 4 50 g 1 50 2 50 4 4 33 50 4 0 — .5 80 15 g o 12 3 00 g .66 12 1 50 2 10 5 00 7 2 3 82 15 g Q q 4 o 00 0 2 .8 0. 83 15 g 12 19 \ 00 g .66 16 2.00 12 12 14 1 i6 1.00 12 0 0 Av g 4 5 3 \ iX) 6 6 1 83 9.3 1 33 g !75 4 6 8.6 3 08 50 6 6 i. 103 15* Q g o 00 A 4 1 .00 o 00 4 2 50 .00 3.2 -2 78 16 g A ft 4 4 1 .00 4 1 00 0 4 2.4 1 88 16 g 16 24 1 50 20 83 12 60 12 1.00 20 12 60 75 16 8 — 1 85 16 B 0 4 0 XO 0 4 2 2 1 00 L6 1. Av. B 5.3 9.3 i 50 8 .41 5 3 80 6 6 iloo 7.3 6 80 2^i 6.9 .3 96 16 Q g g 1 00 12 1 .50 4 10 2 50 7 0 4. 98 16 VJ 20 g 40 4 50 g" 2 00 14 g 43 40 10* 0 —4 106 16 G 8 4 50 4 Too 8 200 4 .50 6 6 1 00 50 5 6 -X. 100 16 G 0 0 4 0 .00 0 0 0 8 0. Av. G 4 50 6 '80 4 .83 6 i'33 5.7 5 1 i 31 .30 5.5 — .2 87* 17 B 6 0 0 4 0 00 0 2 .8 0. 90 17 B 4 0 00 12 4 33 4 1 00 2 4 2 00 i loo 4.8 0. Av. B 2 00 6 4 33 2 50 1 3 2 00 1.00 2 8 0. 95 17 G 0 0 4 4 l.OO 0 00 0 2 1.6 0 105 18 G 0 0 4 4 1 CO 4 1.00 0 4 2.4 — 1. 107 21 G 8 0 00 0 .00 0 4 8 2.4 -2. 101 24 G 0 0 s 0 0 0 .0 0. Ave B 3.4 3.8 .79 6.1 2 40 7.5 1.33 5.8 .62 3.4 6.7 2 60 .96 5.2 .7 Ave G 3 1.5 .41 3.1 1.67 2.5 68 3.7 1.00 2.8 3.2 1 18 98 3.0 - .07 Ave B& G 3.1 2.7 .61 4 7 2.04 5 4 104 4.8 .75 3 1 5.1 1 97 .97 4.2 .35 5. RESULTS (See Tables 13, 14, 15, and 16) We have tabulated the scores in two ways. First, we have given the results in Tables 13 and 14 in terms of a combined quantitative and quali- tative score, allowing 4' ; for a correctly added column and 2% for an incor- rectly added column. Since the test was given as a time or speed test it seems justifiable to give half credit for columns incorrectly added because it probably required just as much time to sum up a column that was incorrectly added as one that was correctly added1, and because the mistakes frequently may be largely due to the necessary hurry of the work. 'Of course, this would not be true of careless pupils who were less concerned about the accuracy of their work, than covering a large number of columns. We eliminated the records of four of the normal pupils who seemed from the internal evidence merely to have written sums at the foot of the columns without actually summing up the figures. One •'dull" boy of 14 in the seventh grade and an "average" boy of 8 in the third grade seemed to have written mere guesses; a ' bright" girl of 13 in the seventh grade seemed from the answers to have added together a number of columns. She wrote such answers as the following: in 2, "2935;" in 3, "3313;" in 4, "4533;" and in 5. "97740." One "dull" girl of 10 in the fifth grade added 3 columns correctly in the first sitting. After that all were wrong. She covered 5 columns in the second sitting, 6 in the third, 32 in the fourth and 24 in the sixth. 60 The Measurement of Mental Traits in Second, in Tables 15 and 16 we have based the scores on the columns correctly added (accuracy scores), allowing no credit whatever for columns incorrectly added. The reader may base his conclusions on either of these two methods of scoring; we shall compare the results of both methods. COMPARATIVE EFFICIENCY Based on the columns both correctly and incorrectly added, Table 13. the average efficiency for all the normals for the five sittings was 19.5' < and for all the epileptics, Table 14, 7.5'.', . The epileptics' average was only 38' < as high as the normals' average. The epileptics did not do quite as well as the youngest group of normals, and only 25% as well as the oldest group. Basing the results only on the columns correctly added, Tables 15 and 16. the corresponding average for the normals is 14.7%, which is 75% of the normals' combined score in Table 13, while for the epileptics it is 4.2' j . which is only 56% of the epileptics' combined score in Table 14. The epi- leptics' comparative efficiency is only 28% of the normals when the scores are based only on the columns correctly added. They did a little better than the normal seven-year-olds, but only 17' , as well as the oldest group. In other words, the epileptics added relatively more columns incorrectly than did the normals. The boys' ability to add, based on the combined averages in Tables 13 and 14, for all the boys in all five sittings was very slightly less than the girls' among the normals, but decidedly higher among the epileptics. The normal boys did 98% as well as the normal girls, while the epileptic girls only did 61% as well as the epileptic boys. The girls did better in three sittings and poorer in two sittings among the normals and poorer in all the sittings among the epileptics, based on the general averages at the foot of the col- umns. Based on the averages for all the sittings the normal girls did better than the normal boys in four ages and poorer in seven ages, while the epileptic girls did better than the epileptic boys in only two ages and poorer in four ages, no comparison being possible in the other ages. W hen the results are based on the corresponding averages for the scores correctly added in Tables 15 and 16, no difference is found between the gen- eral averages of all the five sittings for the normal boys and girls, while the girls are superior in three sittings and inferior in two, based on the general averages at the foot of the columns. On the other hand, the epileptic girls did only 57', ', as well as the epileptic boys in the general averages, while they were decidedly inferior to the boys in all sittings except one. The girls did better than the boys among the normals in six ages and poorer in five. Among the epileptics they did better than the boys in only two ages and Normal and Epileptic School Children 61 poorer in four, while no comparative data are available in the other ages. We find, therefore, no significant difference in this addition test between the normal boys and girls, but the epileptic boys did decidedly better than the epileptic girls. Based on the averages for the two sexes for all the five sittings, the per- formance in this test improves with ascending age among the normals in eight ages while it decreases in only two ages, when the scores in each age are compared with the scores in the next higher age. The results are the same whether based on Table 13 or Table 15. The losses occur in age nine, in which 57' < are backward or dull (which, however, is a smaller per cent than the per cent of backward or dull in age eight, while it is the same as in age ten) ; and in age 16, in which 75% of the pupils are classed as average or dull. The improvement shown in this test with increasing age is in accord- ance with expectation, because the older pupils are not only more mature but have had more training in the use of the fundamental mathematical processes. The difference between the averages in the lowest and in the highest age amounts to 20.9% (24.7% -3.8' A , Table 15). But here, again, the score for age 15, in which there are no dull pupils, is higher than the score for age 17, in which all the pupils were average or dull. There is no consistent increase among the epileptics. The records of the four oldest epileptics are almost on a par with those of the youngest epileptics. Based on the combined scores, one epileptic was not able to score a single point in any sitting, two failed entirely in two sittings, and nine in one sitting. Based on the scores for only the columns correctly added, three failed com- pletely in all five sittings, five in four sittings, four in three sittings, six in two and six in one sitting. Only some of these failures could be attributed to antecedent convulsions — notably in the case of the single failures of Nos. 102. 79. 86, 103. and 95. Among the normal children, only one child (No. 5, again) failed in all the sittings based on the combined scores, while two failed in one sitting, an "average" boy of eight and a "dull" girl of ten, both in the second grade. When the scoring was based only on the columns cor- rectly added, three failed in all the sittings, two in four, two in three, seven in two. and eleven in one. Both the normals and the epileptics in both tables made their highest average in the fourth sitting and their lowest average in the second sitting. COMPARATIVE IMPROVEMENT The average monthly improvement made by each normal subject during all the sittings was only .3% . when the combined scores are considered, Table 13, while there was an average loss of .2% when the scores are based only on the columns correctly added. Table 15. The gains for the epileptics, based on the corresponding scores, were .63'/ and .35%.' Based upon the com- 62 The Measurement of Mental Traits in bined scores, the epileptics gained 210% more than the normals. According to the indices of improvement, the gain in the fifth sitting over the first (/ f) amounted for the normals to 18% when based on the combined scores and only 2% when based only on the columns correctly added. The correspond- ing figures for the epileptics are 35% and a loss of 3%, a difference of 17% in favor of the epileptics in the first instance and of 5% in their disfavor in the latter instance. The gain in the last two sittings over the first two (/ ?:§) amounted for the normals to 22% when based on the combined scores and only 4% when based only on the columns correctly added. The correspond- ing figures for the epileptics are 105% and 4%, a difference, respectively, of 83% and 0% in favor of the epileptics. The following are the per cents of gain based on each index of im- provement: Combined Scores Correct Scores Only Index ...... f | | | J | | 5 Normals 5% 10% 15% -1%> 8% 13% 24% -3% Epileptics 10% 51% 18% 8% 40% 104% 4% -25% Based on the combined scores, the normal pupils gained 15% more than the epileptics in the second sitting, while the epileptics gained 41% more than the normals in the third sitting, 3% more in the fourth, and 9% more in the fifth. The greatest variation from sitting to sitting is among the epileptics. In one sitting they improved 104% while in another they lost 40%, based on the results for the columns correctly added. Some epileptics who did abso- lutely nothing on some days made a fair score on other days. Based on the figures in the column giving the average monthly gain for sittings 1 to 5, combined scores, 10 epileptics lost and five made no improvement, which is 50% of all the epileptics, while the corresponding figures for the normals are 25 and 8, which is 45% of all the normal pupils (72 in this table). Based on the scores for the columns correctly added in Tables 15 and 16 the figures for the epileptics are six losses and fourteen no gains, which is 66% of all the epileptics, and for the normals thirty-two and nineteen, which is 67% of the normal children. The absolute average monthly gain was twice as great for the normal boys as for the normal girls (.4% vs. .2%), when based on the average score for the columns correctly and incorrectly added, while the loss for the boys was three times as great as the loss for the girls when based only on the scores for the columns correctly added (-.3% vs. -.1%). Among the epi- leptics the gain was over six times as great for the boys as for the girls, based on the figures in either table. Based on the average indices of improvement at the bottom of the columns, the normal boys gained more (or lost less) than Normal and Epileptic School Children 63 the normals girls in three successive indices, and gained less in three, in both tables, while the epileptic boys gained more than the epileptic girls in all indices except one, for the combined scores and except two for the correct scores. The gain in the fifth sitting over the first (/?), combined scores, was only 2% greater for the normal boys than for the normal girls (19% vs. 17%), but 61 % greater for the epileptic boys than the epileptic girls (63% vs. 2% ). Based on the scores for the columns correctly added, the gain was 14' < greater for the normal girls than for the normal boys (-5% vs. +9%), while the loss was 2% less for the epileptic girls than for the epileptic boys (-2', vs. -4', i. The gain based on the average of the last two sittings compared with the average of the first two was 8% greater for the normal girls than for the normal boys (26% vs. 18%), and 65% greater for the epileptic boys than for the epileptic girls (134% vs. 69%), when the re- sults are based on the combined scores. When the results are based only on the columns correctly added, the normal boys gained 2% more than the normal girls (o( , vs. 3%), and the epileptic boys 142% more than the epi- leptic girls (160% vs. 18%). It is evident that the results are not con- sistent. Apparently the girls, both normal and epileptic, do comparatively better when the results are based only on the columns correctly added, but the epileptic boys tend rather consistently to improve more than the epileptic girls. There is a very slight tendency for the improvement among the normal pupils to increase with ascending chronological age. When the average monthly gains (^5") are compared in each successive age (as indicated on p. 50), there is improvement in six of the successive ages and losses in four, in both tables. When the results are based on the index of \L\t there is an increase in seven ascending ages and a decrease (or equal scores) in only three, based on the figures in both tables. The results for the epileptics are too irregular to justify comparison. 6. CONCLUSIONS 1. The ability of our normal pupils to add as determined by this test was decidedly inferior to the ability shown in the same test under the same time limitations by our dental group of children, the dental group doing almost 100% better (37.6% vs. 19.5% ). We do not know how to account for this large difference. It might be assumed that it is due to the larger number of younger children in our New Jersey group than in our Cleveland group. But the Cleveland children are superior in every age from 10 to 16 *here comparisons are possible. The explanation may be that better train- ing was afforded in arithmetic in the Cleveland schools. It should be re- membered, however, that the Cleveland average is based on the six series of tests given in the course of a whole year. 64 The Measurement of Mental Traits in Based on the results of these two groups of normal children we may say that this test, under the conditions in which it is given, is applicable for use with normal school children over 7 or 8 years of age. 2. The epileptics' ability to add was decidedly inferior to the ability of normal children to add, and this is particularly true so far as concerns the epileptics' ability to add correctly. The epileptics did only 38% as well as the normals on the combined scores and only 28 % as well on the accuracy scores. 3. There is no significant difference in the ability to add between the normal boys and girls in this experiment. The odds, however, are slightly in favor of the girls. In the previous use of the test with the dental group of normal children, the boys did decidedly better. Among the epileptics in this experiment the boys excelled decidedly. 4. The capacity to add increases from age to age among normal chil- dren, due to increasing mental maturity as well as increasing practice. The oldest pupils do over three and a half times as well as the youngest, based on the combined scores, and six and one-half times as well, based on the accuracy scores. 5. Large individual differences exist in the ability to add, as shown by the extreme variations in the individual scores in each sitting, based (a) on the combined scores (Tables 13 and 14) and (b) on the columns correctly added (Tables 15 and 16). (a) (b) (a) (b) (a) (b) For the epileptics 0 to 22'/! 0 to 20% 0 to 24% 0 to 24% 0 to 20%. 0 to 20% For the normals 0 to 58',' 0 to 56% 0 to 70% 0 to 64% 0 to 74% 0 to 72% (a) 4 .bi (a) .5 ib) 0 to 20% 0 to 20% 0 to 24% 0 to 20% For the normals, .... 0 to 78% 0 to 76% 0 to 76% 0 to 76% The individual differences in each age for the normals are as follows, based on the averages for all sittings (1-5) : Age 7 8 9 10 11 12 13 14 15 16 17 Difference (a) — 12.0 12.0 13.0 14.* 23.6 14.9 29.6 54.2 21.2 13.8 30.4 Difference (b).. 5.6 15.2 10.6 20.8 24.8 14.8 44.0 67.8 30.0 22.6 37.6 5. The average monthly improvement during the five sittings was, on the whole, greater for the epileptics than for the normal children, both abso* lutely and relatively, whether based on the combined scores or only on the accuracy scores. The greater improvement made by the epileptics is due to their poor initial records and their great variability, which produced oe- Normal and Epileptic School Children 65 casional excessive gains. The monthly improvement in absolute terms, however, was small for either group. One-half of the epileptics as against 45' t of the normals made no improvement or actually lost during the experiment. 7. We do not find any significant difference in the average amount of monthly improvement between the normal boys and girls, although in our previous experiment with the dental group the boys improved more than the girls. The epileptic boys clearly improved more than the epileptic igirls. 8. There is a very slight tendency toward an increase in the monthly improvement as the normal children grow older. 9. The increase in ability to add with increasing age and with time indicates that this test can be used to measure growth in mental capacity, resulting from increasing maturity and from practice. Simpson found a correlation in an addition test of .72 with the estimated intelligence in his "good" group of subjects.1 Chapman found a higher correlation (0.96) between his initial and final scores in his addition test than in any of his four other tests.2 We would emphasize, however, that the test presupposes specific training or familiarity with the process of addition; success in the test depends, of course, on other factors than general intelligence. 'Benjamin R. Simpson, Correlation of Mental Abilities, 1912. -J. Crosby Chapman. Individual Differences in. Ability and Improvement and their Correlations, 1914. We cannot here attempt to assemble the considerable experimental literature which exists on some of the tests which resemble our own The literature has been summarized by Guy Mr ntrose Whipple in his Manual of Mental and Physical Tests: Part I, 1914, and Part II, 1921. M T 5 V. CONTROLLED ASSOCIATION Antonym Test With Supplied Key-Words: 1)1 to D5 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS Twenty-five words were printed in 12 point type on each of six blanks, measuring 4x12 inches. These blanks also served as record sheets. The 150 words, used in the complete set of blanks (only the first five of which were used in this experiment) were selected from a larger preliminary list. The attempt was made to select 150 simple words, all of which should, if possible, be familiar to even younger children and all of which should have familiar antonyms. It is no easy task to arrange six lists containing 25 words each of equal difficulty. In the first place, it is difficult, if not impossible, to select 150 words the meanings of which are equally familiar to children of different ages. Second, it is still more difficult to select 150 words with equally familiar antonyms. The opposites of some words are suggested very fre- quently by our daily experiences, while the opposites of other words, the meanings of which may be equally familiar, are rarely suggested. Some words allow of several synonymous antonyms, or antonyms corresponding to the various shades of meaning attaching to the words, while others have only one antonym. Some words permit one or several partially correct antonyms, while others have only one recognized opposite. In view of these difficulties, which would have led us to reject this test had it not been for the recognized merit of the opposites test as a test of intelligence, it seemed to us that the only feasible way of constructing approximately equal lists of words was to distribute in each test an approximately equal number of easy and hard words — i. e., easy and hard from the standpoint of the familiarity of their antonyms — to allow one-half credit for partly correct antonyms, and to in- struct the subjects to pass over difficult words, but to return to them later if the time permitted.' We distributed the 150 words, as best we could, according to this plan and then submitted the six lists to two instructors in a teachers' training school, with the request that they rearrange them inde- pendently of each other, so that, according to their judgment, each test would contain an equal number of easy and difficult words, and so that the six lists would be of equal difficulty. The final arrangement represents the best compromise that could be made from the judgments of three workers. It would, of course, require an extended investigation on children of different ages experimentally to construct such lists as these, or to demonstrate whether or not our lists are of equal difficulty. In our previous experimental use of 'A number of subjects followed this suggestion, but we do not know how many were able to return to difficult words. Judging by the records very few had time to do so. 66 Normal and Epileptic School Children 67 this test there was a consistent and considerable increase in the scores from sitting to sitting.1 In the present investigation, however, the increase was not uniform for either the normals or the epileptics, while the highest score was made by the normals in the fourth sitting and by the epileptics in the third sitting. In both investigations the lowest score was made in the first sitting, due probably to lack of familiarity with the test. Our general impression of the lists is that they are fairly satisfactory, although not absolutely uniform in difficulty. 2. NATURE AND PURPOSE OF THE TEST This is recognized as one of the best tests of intelligence of the so-called controlled association or selective thinking type. Simpson found that the degree of correlation between his easy opposites test and the intelligence of people in general amounted to 0.82, while for the hard opposites test it reached 0.96. 2 In order properly to perform the test the subject must possess ac- curate knowledge of the meanings of the key-words, he must be able to con- ceive of the opposite meanings of the words, to do which he must be able to compare, and particularly to abstract differences, and, finally, he must be able to select the right word to express the difference. The number of words written in the test represents the speed of associating antonyms with the key- words plus the time required to write the antonyms. As we have stated in Chapter III (which see) the necessity for writing the associates delayed and complicated the responses. It is possible that the older and more intelligent subjects mentally "filled in" more antonyms than they had time to write, while the younger children probably were embarrassed by the difficulty of writing and spelling. 3. DIRECTIONS FOR GIVING THE TEST General Directions: See pages 15 to 17. Order of Giving Tests: First sitting, Dl. Second sitting, D2. Etc. Time: 60 seconds (one minute, precisely). Place the papers upside down on the pupils' desks. Caution pupils not to turn them until directed to do so. Instructions to Pupils: "On the other side of the papers placed on your desks is a list of words. The moment I say 'turn' you must turn your papers over at once, and write opposite each word the word which means just the opposite; that is, has just the opposite meaning. For example (explain, using these illustrations on the board): Better — worse. Correct — incorrect. Sunset — sunrise. If you cannot think of the opposite of any given word pro- ceed to the next, and return to the word later if there is time. Write as many 'Dental Cosmos, 1912, 404f. and 545f. ^Benjamin R. Simpson. Correlation of Mental Abilities, 1912. 68 The Measurement of Mental Traits in correct opposites as you can. If you are troubled about the spelling of words, spell them as well as you can, and don't lose any time trying to think of how to spell words. Nothing will be taken off for words spelled incorrectly." (During the later sittings the first four or five words of Dl may be used as illustrations.) "Now, ready!" "Turn!" "Now, stop!" "Turn!" "Sign your names." Scoring: The grading in this test is based on the quality of the re- sponses. Each perfect antonym, 4%. Each partly correct antonym, 2%. Correct idea, but expressed in two words. 2% (but if a conventional phrase, 4', ). No credit when the key-words are repeated and prefixed by "not." For errors in the form of the word (which are not mere errors in spelling the word) deduct 2%. The following "key" prepared largely on the basis of the words which have been actually given in the experiments thus far conducted with the test, was followed in grading the papers: KEY TO GRADING OF TESTS Dl TO D6 Dl Perfect 4% Partly Correct 2% Someone Noone, Nobody, None Idle Industrious, Busy, Occupied, Employed Ambitious, Energetic, Working, Thrifty Useless Useful, Usable Serviceable, Beneficial, Worth, Something of use Gain Loss, Deficit, Lose Privation, Waste, Damage Naked Clothed, Clad, Robed, Gowned, Covered, Dressed Ugly Beautiful, Pretty, Attractiveness, Splendid, Ornate, Good looking, Good Handsome, Nice Odd i Even, Ordinary, Usual, Common, Familiar Rim Center, Hub Inside, Middle Wider Narrower Closer, Thinner Break Mend, Repair, Join, Rebuild Construct, Make, Fixed, Fasten False True, Genuine, Right, Correct Honest, Ingenuous, Good, Real Purity Impurity, Uncleanness, Corruption, Impure, Foulness, Uncleanliness, Unclean, Dirty Dirtiness Wife Husband Man Cloudy Clear, Bright, Fair Sunny, Moonlight, Starry, Sunshine, Shiny Top Bottom Deep Shallow, Superficial, High Bachelor Maid, Spinster, Virgin, Married man Married Front Back, Rear Behind, End Ripe Green, Unripe, Raw, Undeveloped Immature Normal and Epileptic School Children 69 Teacher Pupil, Scholar, Learner, Disciple, Student Large Small, Little Spend Save, Horde, Accumulate, Retain Empty Full, Fill, Filled Ceiling Floor Hilly Flat, Plain, Level D2 Perfect 4% Bad Good, Virtuous, Right Outside Inside, Internal Scarce Plentiful, Abundant, Frequent, Common Quick Slow, Sluggish Shaky Steady, Firm Known Unknown, Unfamiliar Big Little, Small Seldom Frequently, Often, Frequent Less More Buyer Seller, Salesman, Vender Sugar Salt Doubtful Certain, Sure, Doubtless Weak Strong, Mighty, Vigorous, Robust, Powerful White Black Stale Fresh Like Dislike, Unlike, Different, Dissimilar Birth Death Either Neither Sick Well, Healthy Glad Sad, Sorry, Unhappy Thin Thick Brother Sister Many Few Above Below, Beneath, Underneath Sink Rise, Ascend, Mount, Soar, Float D3 Perfect 4% Wrong Right, Correct, Proper Up Down Wet Dry New Old High Low Cool Warm Inconsiderable, Insignificant Hold, Keep Replenished, Replete Base, Foundation Smooth, Even Partly Correct 2% Excellent, Moral, Correct In, Inner, Within, In it, Indoors General, Many, Plenty, Enough Lazy, Slowly Established, Strong, Even, Fast, Still Unrecognized Short Always, Every time Increased Hawker, Peddler Vinegar, Acid, Sour Know Able, Capable, Efficient Light, Brown Preserved, Wholesome Hate, Don't like Decay None Healthful, Wholesome, Comfortable Mad, Angry, Sorrowful Fat, Bulky, Stout None, Infrequent, Rare Under Swim Partly Correct 2% Below Aged, Ancient Deep Hot, Heated 70 The Measurement of Mental Traits in Soft Hard Laugh Cry Tight Lcose, Slack Always Never Happy Sad, Unhappy, Sorry Wise Unwise, Foolish, Stupid, Imprudent, Indiscreet, Silly Follower Leader Forget Remember, Retain, Hold Winter Summer Blunt Sharp, Pointed, Acute Strife Peace, Concord, Amitv Off On Wild Tame, Civilized, Cultivated, Peaceful, Domesticated, Sane, Gentle Beginning End, Terminus, Close, Conclusion, Ending, Finishing, Finish- Coarse Fine, Refined, Polished, Cultured, Cultivated Something Nothing Liquid Solid Push Pull, Draw, Drag Queen King D4 Perfect 4% Great Little, Small Hot Cold Dirty Clean Able Unable, Impotent, Incapa- able. Weak, Inefficient, Ineffective, Incompetent, Feeble Late On or in time, Punctual, Early Miss Mr., Master, Hit, Strike, Catch Dress Undress, Disrobe, Suit of Clothes (or Coat and Pants) Sowing Reaping, Harvesting, Mowing Left Right, Came, Taken Height Depth, Lowness This That Yes No, Nay Dark Light, Fair, Bright, Clear Evil Good, Beneficial Cowardly Brave, Bravely, Courage- ous, Daring, Heroic, Bold, Valiant, Fearless Stony, Rigid, Unyielding, Rough Frown, Moan, Cried No time Sorrow, Grieved, Sorrowful Ignorant, Dull, Dumb Remind, Know Polished Love Calm Smooth, Thin Dry Ruler Partly Correct 2% Insignificant, Inferior Cool Pure Can't, Cannot Man, Mrs. Coat, Pants Picking Low (Length? Width?) Breadth White Wholesome, Right Manly Normal and Epileptic School Children 71 North South Yours Mine Drop Hold, Pick up Open Closed, Shut, Ajar, Close Round Square War Peace, Amity, Concord East West Raise Lower, Depress, Lay down, Depreciate Rough Smooth, Sleek, Refined, Calm, Tranquil Noisy Silent, Quiet, Still D5 Perfect 4% Day Night Asleep Awake Heavy Light Best Poorest, Worst Add Subtract, Detract, Diminish, Take away Careless Careful, Cautious, Heedful, Circumspect Same Different, Opposite Loud Soft. Low, Softly, Quiet Expensive Cheap, Economical Giving Taking, Receiving Joy Sorrow, Grief, Misery, Melancholy, Sadness Much Little Near Far, Distant, Remote Profit Loss, Debit Come Go, Leave Woman Man Tall Short, Stunted Absent Present, Here Son Daughter Country City, Towr. Straight Curved, Crooked Backward Forward, Frontward Nowhere Somewhere Success Failure Lean Fat, Stout, Plump D6 Perfect 4% Girl Boy, Lad In Out Alike Different, Unlike, Dissimilar Mind Body, Disobey, Disregard Future Past My Raise, Lift, Keep Locked Lean, Flat, Straight Drop Even, Easy, Fair Partly Correct 2% Wake Animated Decrease Anxious, Thoughtful, Neat Unlike, Another Silent, Weak, Still Keeping, Stingy Less Away Waste, Damage Little, Low, Small Round, Bent Everywhere Thick, Heavy Partly Correct 2% Outside Opposite Ancient, Old, Now, Present 72 The Measurement of Mental Traits in Him Her Over Under, Underneath Below Slowness Fastness, Rapidity, Quick- ness, Promptness, Speed Sweet Sour, Acid, Bitter, Salt Foot Head, Hand, Top First Last, Hindmost Morning Evening, Eve Night Broad Narrow Thin Dead Alive, Living Animate Ocean Land Continent Hollow Solid, Filled, Genuine Rooster Hen Chicken Upper Lower Here There Away Love Hate, Hatred, Malice, Dislike Coldness Sharp Blunt, Dull, Obtuse Rich Poor Young Old, Aged Stay Go, Depart, Leave, Proceed Friend Foe, Enemy, Adversary, Antagonist Normal and Epileptic School Children 73 4. TABLES TABLE 17 Antonym Test: Dl— 5 Normal Children No. Ae. Sx 2 8 f\ 4 % fi f 9 4- 7 5 m f\ 1- 91 5 Gain Ave. % 2 7 B 2 0 14 0 7 00 12 0 85 10 0 B3 12 0 net 1-20 8 0 . . 11 i 37 6 00 10 0 2 D o L 7 G 6 0 20 0 1 66 30 0 2 .00 22 0 BO 12 0 CQ , 3 J 13 0 17 U i 25 1 32 17 2 1 . / 4 7 Av . 4 0 16 0 4 33 21 0 1 .42 16 0 .81 12 0 .75 10 5 14 0 i 34 3 66 13 6 2.1 8 8 B 6 7 13 7 2 41 24 5 1 69 20 0 63 14 5 .64 10 2 [€ 0 i 45 3 01 15 4 1.9 3 8 G 1 3 6 6 SO 5 3 1 .06 26 0 50 0 .00 4 0 1 a 38 00 3 2 — .3 11 8 Av. 5 2 11 8 2 05 19 2 1 59 14 8 60 10 5 .50 8 5 12 0 i 09 2 41 12 0 1.3 3 9 B 8 0 9 0 2 25 22 6 1 89 18 6 .57 8 0 .42 12 6 a a i 05 1 00 14 3 0. 4 9 G A U 03 3C OO 3 o Q j LO n u i i 03 z D ■ 57 15 E 24 0 } 44 1 1 OO 21 0 1.6 7 9 Av 10 8 12 0 1 44 29 0 1 90 24 0 83 15 l .52 14 3 5 i 31 1 32 18 1 .9 4 10 B 21 0 33 3 2 20 43 3 1 28 41 0 i 03 40 0 .98 27 7 40 5 i 58 1 95 35 1 5 4 o O 10 G 21 3 29 0 2 29 36 6 1 31 35 3 94 30 6 .09 25 3 OJ A U i 62 3 59 30 6 2.3 7 10 Av 21 1 31 3 2 25 40 0 1 30 38 5 99 36 0 .94 26 37 4 i 60 2 65 33 2 4 .0 3 11 B 25 3 32 6 1 38 49 3 1 91 44 0 94 46 6 1.03 29 0 45 3 i 55 1 72 39 6 5.3 8 11 G 26 7 34 2 1 23 46 7 1 48 43 6 87 37 7 1 00 30 5 42 3 i 43 1 53 38 3 3.6 11 11 Av 26 3 33 1 •1 27 47 4 1 60 43 7 83 40 4 1.01 30 0 43 1 i 46 1 59 38 6 4.0 5 12 B 36 0 42 8 96 44 0 1 02 52 5 2 78 42 4 .80 39 4 45 2 i 01 1 14 42 7 1.6 3 12 G 22 0 34 6 2 87 50 0 1 35 51 3 1 16 42 0 .74 28 3 46 0 i 63 2 33 40 0 5.0 8 12 Av 30 7 39 7 1 91 46 2 1 16 52 0 2 08 42 2 .78 35 2 45 7 i 28 1 65 41 7 2.7 4 13 B 44 0 52 0 1 13 55 0 1 16 53 5 99 48 0 88 46 2 50 7 i 06 2 57 49 g 1 0 2 13 G 33 0 27 0 57 52 0 1 94 32 0 53 46 0 1.91 30 0 39 0 i 70 1 00 38 0 3^2 6 13 Av 40 3 42 0 99 54 0 1 42 46 3 83 47 3 1.19 40 5 47 0 i 27 2 25 45 8 1 8 4 14 B 46 5 48 0 1 39 46 0 88 63 3 1 07 47 3 3 26 47 2 49 2 i 00 42 48 6 1.5 7 14 G 48 8 54 5 1 23 66 0 1 18 63 1 95 62 2 1 00 51 7 62 7 i 25 ! 46 58 1 3.6 11 14 Av 48 0 52 1 1 29 57 0 1 05 63 2 1 09 57 8 1.68 50 0 57 8 i 16 i 44 54 6 2.9 2 15 B 88 0 91 0 1 03 80 0 81 90 0 1 10 94 5 1.04 89 5 87 2 98 i 07 89 3 1.8 2 15 G 82 0 77 0 94 79 0 1 02 85 0 1 07 93 0 1.09 79 5 89 0 i 11 i 13 83 2 2.7 4 15 Av. 85 0 B4 0 98 79 3 95 87 5 1 08 93 7 107 84 5 88 1 i 06 i 10 86 2 2.3 1 16 B 66 0 46 0 69 62 0 1 34 12 0 83 56 0 57 0 i 01 78 56 5 -4.6 3 16 G 63 3 62 0 98 66 6 1 08 56 0 81 64 6 1.14 62 3 60 3 95 i 01 61 6 .3 4 16 Av 64 0 56 6 8b 66 6 1 08 57 5 94 61 5 1.06 60 6 59 5 96 95 60 3 - .7 1 17 B 56 0 42 0 75 60 0 1 42 54 0 90 64 0 1.18 49 0 59 0 i 20 i 14 55 2 2. 2 17 G 70 0 56 e 1 07 76 0 1 04 74 0 1 01 88 0 1.19 73 0 81 2 i 18 i 32 76 3 5.2 3 17 Av 65 3 49 0 91 70 6 1 17 67 3 97 81! 3 1.19 65 0 73 8 i 19 i 26 69 2 4.0 Ave. B 29 2 34 8 1 72 39 4 1 37 41 8 1 11 36 5 1.06 31 9 39 1 i 24 i 92 36 2 2.1 Ave . G 33 37 6 1 41 4> 8 1 37 45 2 92 44 0 .92 35 5 44 6 i 29 i 62 41 8 2.7 Ave. B & G 31 5 36 2 1 55 44 2 1 37 43 6 1 01 40 3 .98 33 7 42 0 i 27 i 77 39 1 2.4 74 The Measurement of Mental Traits in TABLE 18 Antonym Test: Dl— 5 Epileptics Su . Ae. Sx. k t / _ 1 /f 4 5 1 "2 - 4~o % M fi l-5 % Gain Ave. V 81 10 B 0 4 0 .00 o. 2 1 3 0. 102 10 G 0 '6 6 6 12 2.00 0. 9 3.6 3 104 10 G 0 0 0 4 0 .00 2 9 0. Av. G 0 0 0 5 6 1.00 0. 5 5 2 7 15 99 11 G 0 0 2 0 .00 0. 0 .5 - 0. 79 12 B 0 0 0 4 0. 2 10 1.3 84 12 B 12 16 1 33 32 2 00 24 .75 24 LOO 14. 28 o 00 2.00 21.6 3. 91 12 B 0 0 4 0 .00 0 0. 0 8 0. Av. B 4 5 3 i 33 12 2^00 9.3 .37 12 1.00 4.6 10 2 00 2.00 8 2 13 ios 12 G 0 0 0 0 0 0 0 0. 93 13 B 0 0 0 0 0 0 0 0. 0 80 13 B 0 0 0 0 0 0 0 0. 0. 97 13 G 0 20 8 .40 4 50 10 4 40 8 1.3 77 14 B 8 24 3 00 38 1.58 18 .47 24 1 33 16 21 1 31 3 66 22 4 4 89 14 B 6 12 2 00 0 .00 8 20 2.50 9 14 1 00 3 33 9 2 3.5 Av. B 7 18 2 50 19 .79 13 .47 22 1.91 12.5 17. 5 43 3 16 15 8 3.7 80 15 B 0 14 24 1.71 8 .33 8 1 00 7 8 i 14 10.8 2 82 15 B 14 24 i 7) 34 1.41 20 .58 16 .80 19 18 94 in 21 6 .5 83 15 B 4 16 4 00 24 1 50 12 50 12 1 00 10 12 " i 20 3. 00 13.6 2. Air rt V R 0 1 8 Q z fit; 27 .3 1 .04 13 3 . 4/ 1 9 yj 0 c o 0 IC 0 i 09 C VI ICQ ID J i03 15 G 0 0 0 0 0 0 0 0 0. 78 16 B 4 8 2 no 8 LOO 4 ".50 10 2^50 6 7 i 16 2.50 6 8 15 88 16 B 8 20 2 50 20 1.00 8 .40 12 1.50 14 10 71 1 50 13 6 1. 85 16 B 0 20 8 40 12 1.50 18 1.50 10 15 i 50 11 6 4.5 Av. B 4 16 2 25 12 .80 8 .80 13.3 1.83 10 10 6 i 12 2 66 10.6 2 3 96 16 G 0 4 0 .00 0 2 0 00 1 0 0. 98 16 G 0 4 18" 450 12 ;66 .... 15 8.5 4 106 16 G 0 "l 4 200 0 .00 2 1 i 6b 1.6 5 100 16 G 0 0 8 4 .50 0 !6o 0 2 2.4 0. Av. G 0 2 4 i!66 7.3 1.66 35 33 1 4.5 50 3.2 1 87 17 B 0 0 0 0 0 0 0 0. 90 17 B 0 0 0 0 0 0 0 0 0 95 17 G 0 0 0 0 0 0 0 0 0 105 18 G 0 0 0 0 0 0 0 0 0. 107 21 G 4 8 2 00 8 L00 12 L50 16 i 33 6 14 2 33 400 9.6 3 101 24 G 0 4 0 .00 0 0 2 0 00 9 0 Ave . B 3.5 10.2 2 29 12.8 1.17 7.6 .55 10 2 1.31 6 7 8 8 1 28 2 35 8 7 1.5 Ave . G 2 3 1 2 00 2.4 .68 4.0 1.62 3 2 62 1.6 3 6 74 4 00 2.5 .8 Ave. B & G 2.0 7.1 2 25 7.7 1 00 6.1 .88 6.4 1.03 4 4 6.4 1 09 2 56 5 9 1.2 5. RESULTS (See Tables 16 and 17) COMPARATIVE EFFICIENCY The average efficiency for all the normals in all the five sittings (Column 1-5) was 39.1%, and for the epileptics 5.9%. The epileptics were only 15% as efficient as the normal children in supplying antonyms. The epi- leptics did less than half as well as the youngest normals, the seven-year-olds, and about one-twelfth as well as the oldest normals. The boys' capacity, based on the general average at the bottom of the column for all five sittings (1-5). was appreciably inferior to the girls' among the normals but decidedly superior to the girls among the epileptics. The normal boys did 86% as well as the normal girls, while the epileptic girls did only 28% as well as the epileptic boys. The normal girls did consistently better in every sitting than the normal boys, while the epileptic boys did de- Normal and Epileptic School Children 75 cidedly better than the epileptic girls in every sitting, based on the sex aver- ages at the bottom of the columns. Based on the sex averages in the different ages for sittings 1 to 5 (next to the last column in the tables), the normal girls did better than the normal boys in five ages, but poorer in 6, while the epileptic girls did poorer in three ages, and better in two ages than the epileptic boys, the averages were the same in one age, and no data were avail- able in the other ages. The girls in this test proved to be slightly superior to the boys among the normals, but decidedly inferior among the epileptics. There is a very considerable improvement from age to age in this test. Based on the averages for the two sexes for sittings 1 to 5, there is an im- provement in eight ascending ages and a loss in only two, when the averages in a given age are compared with the averages in the next higher age. The losses occur in ages 8 and 16. in which a preponderance of the children were classed as average, namely, 72% in age 8 and 75% in age 16. The decidedly best record was made by the 15-year-olds, three-fifths of whom were classed as bright. The 17-year-olds, who were all classed as dull and average, did only 80% as well as the 15-year-olds. The difference between the youngest and the oldest subjects tested amounts to 55.6%. The records for the epileptics do not show any consistent gain with in- creasing age. Most of the oldest epileptics do worse than the youngest ones. Xine epileptics failed to make a score in any sitting in this test, while fourteen made no scores in one or more sittings. Clearly this test was too difficult for the epileptics. On the other hand, one of the normal pupils failed completely in all of the sittings (Xo. 5, the "dull" and inattentive 8-year-old boy in the second grade), two failed in four sittings (one an "aver- age" boy of eight in the second grade and one a "bright" boy of ten in the second grade), three failed in three sittings, three in two sittings and six in one sitting. Only three of these 15 pupils were above the third grade, while six were in the second grade. Eight of these pupils were eight years old and one seven, and only two were above ten years. A considerable number of seven- and eight-year-old children will fail in this test. Both the normals and the epileptics made their highest score in the third sitting, and their lowest score in the first sitting. COMPARATIVE IMPROVEMENT The average monthly gain made by each subject during all the sittings (last column in the tables) amounted, in the absolute units, to 2.4% for the normals, and 1.2% for the epileptics, who therefore improved only 50% as much as the normals. In relative amount of improvement, i. e., relative to the size of the average scores, the epileptics did better, owing to the fact that their scores were considerably lower than those of the normals. This is 76 The Measurement of Mental Traits in shown by a comparison of the indices of improvement. The percentage of improvement made in the second sitting compared with the first, in the third compared with the second, etc., is shown by the following figures: The normals gained 37% more than the epileptics in the third sitting, 13% more in the fourth, and 18% more in the average of the fourth and fifth compared with the average of the first and second. The epileptics, on the other hand, gained 70% more than the normals in the second sitting, 5% more in the fifth, and 79% more in the fifth sitting compared with the first. In the aggregate the epileptics gained rather more than the normals. The epileptics' records, however, are very erratic, as may be seen from the table, and the improvement is sometimes more apparent than real, due to the fact that a child may fail in one sitting adequately to comprehend the meaining of the test, while in the next sitting he understands just what is wanted. A large improvement under such conditions does riot indicate any increase in capacity, but merely a correct understanding of the problem. In our earlier work on the dental group we had reasons to suspect that the large gains made by a few pupils were due to this circumstance. The normals' gain con- sistently declined after the second sitting, while the relative improvement for the epileptics was higher in the fifth than in the third or fourth sittings. The normals lost in one of the five sittings, while the epileptics lost in one and failed to improve in one sitting. Based on the average monthly gain for sittings 1 to 5, fifteen epileptics, or one-half of the epileptics, made no im- provement, while nine normals showed losses and fourteen made no gains, which is 30% of the normals. The absolute average monthly gain for all the sittings was almost one and a third times greater for the normal girls than for the normal boys (2.7% vs. 2.1%), but almost two times greater for the epileptic boys than for the epileptic girls (1.5% vs. .8%). Based on the average indices of im- provement at the bottom of the columns, the normal boys gained more than the normal girls in all the indices, except ?:f (the indices are equal in f), while the epileptic boys also gained more than the epileptic girls, except in one index. The gain in the fifth sitting, compared with the first, was 30% more for the normal boys than for the normal girls (92% vs. 62% ), while for the epileptics it was 165% greater for the girls than for the boys (300% vs. 135%). The gain based on the average of the last two sittings com- pared with the average of the second and third (/£f)> was «*% less for the normal boys than for the normal girls (24% vs. 29%), while it was 54% more for the epileptic boys than for the epileptic girls (a gain of 28% vs. a loss of 26%). Normals . Epileptics 55% 37% 1% -2*% 77% 125% 0% -12% 3% 156% 27% 9% Normal and Epileptic School Children 77 When each age is compared with the next succeeding age for the normal pupils the amount of the improvement, based on the average monthly gain in the final column, decreases in six ascending ages and increases in only 3, while the gains are equal in one age. Based on the index there is a de- crease in seven ascending ages and an increase in only three. In other words, the older normal children were functioning nearer their limit at the beginning than the younger children. With the epileptics the results are very irregular. 6. CONCLUSIONS 1. In this type of controlled association test the normal children did only about three-fourths as well as the dental group of children in our earlier experiment. This large difference may be due to the fact that the dental group was allowed 25 seconds longer in which to do the test, and also to the absence of children under ten in the dental- group. The dental group was superior in four and inferior in three of the ages that could be compared. It should be stated, again, that the averages for the dental group are based on six tests given during the course of one year. Tests on our two normal groups indicate that the opposites test with our arrangement of words and time limitations can be successfully used with children of eight or nine and over. 2. The epileptics were surprisingly deficient in the ability to give antonyms under the conditions of this test, doing only about one-seventh as well as the normal children. 3. The normal girls slightly excelled the boys in this experiment, but the boys were appreciably better than the girls in the dental group. The epileptic boys were decidedly superior to the epileptic girls. 4. The knowledge of antonym equivalents or the ability to associate words of opposite meaning to given key-words, clearly increases from age to age among normal children, due, no doubt, to increasing experience as well as to increasing mental maturity. Our oldest group of normal pupils does more than five times as well as the youngest group. 5. The individual differences in the ability to give antonyms are strik- ingly large. The following are the extreme variations between the lowest and the highest individual scores in each sitting: 1 2 3 4 5 For the epileptics 0 to 14% 0 to 24% 0 to 38% 0 to 24% 0 to 24% For the normals 0 to 94% 0 to 98% 0 to 96% 0 to 98% 0 to 98% The individual differences in each age, based on the average scores for all the sittings (1-5), among the normal children are as follows: Age 7 8 9 10 11 12 13 14 15 16 17 Difference 22.4 31.8 36.4 25.2 43.0 49.2 46.0 72.7 13.4 36.7 35.0 78 The Measurement of Mental Traits in 6. The improvement shown in this test during the course of the experi- ment does not always indicate an increase in mental capacity, but merely a clearer comprehension of the problem. The large gains made by a few of our normal and epileptic subjects can be thus explained. On the whole, the normal pupils improved more than the epileptics. They improved twice as much in absolute units while the greater relative gains (based on the indices) made by the epileptics in some sittings are due to very inferior initial scores, or to abnormally large gains, due to the reason given above, or to the fre- quent inability to compute the indices for many subjects because of the presence of zero scores, which circumstance unavoidably introduces errors. One-half of the epileptics did not make any gains as against 30% of the normal pupils. 7. Although the results are somewhat discrepant, the average monthly improvement in this test is slightly greater for the normal girls than for the normal boys. In the experiment on the dental group the girls likewise im- proved more than the boys. But the epileptic boys, on the whole, gained more than the epileptic girls. However, in view of the discrepancies we cannot conclude that the sex differences are significant in this test. 8. There is perhaps a slight decrease in the average amount of monthly improvement with ascending age, which would seem to indicate that the older pupils were functioning nearer their maximum at the outset — and that some of the younger pupils made abnormally large gains because of a lack of clear comprehension of the test at the beginning. 9. This appears to us to be one of the most valuable tests for measur- ing developing intelligence, as indicated by the consistent increase in the age curve among normal pupils. But as a test for measuring improvement in mental growth with lapse of time it is subject to the weaknesses pointed out above. Chapman also found great irregularities in the opposites test and a lower degree of correlation (.59) between the initial and final scores than he found in any of the other four tests which he used.1 Proficiency in the test, of course, is dependent upon familiarity with words and their opposite mean- ings, and this is dependent not only upon memoriter knowledge of words, but upon the ability to compare, analyze, and abstract constrasting aspects of our experience. For clinical purposes the test is probably best given orally instead of in writing. ]J. Crosby Chapman, Individual Differences in Ability and Improvement and their Correlations, 1914, p. 31. VI. ATTENTION AND PERCEPTION Speed and Accuracy of Perceptual Discrimination in the Crossing Out of A's: El to E5 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS Twenty-six lines of capitals were printed in 12 point type on sheets of paper measuring 6T-)xl2 inches. A test sheet is reproduced below. The capitals were distributed promiscuously in each line. No attempt was made to make all the lines exactly equal in length. On the other hand, in order to secure successive sheets of equal difficulty great care was taken to change the placement of the letters in the different lines, so that no duplicate lines would occur on the different sheets, particularly in the lines that the subjects would probably have time to get over (and only a few such duplica- tions occur), and so that there would be approximately the same number of A s (the letter which was to be crossed out) in each successive group of four or five lines. The following tabulation shows that approximately the same number of A's occurred in the first five, ten, fifteen and twenty lines in each of the successive sheets: Number of A's Number of Test 1 2 3 4 5 6 In first 5 lines 39 37 38 37 36 40 In first 10 lines 78 77 74 77 78 81 In first 15 lines 116 113 116 115 116 115 In first 20 lines 155 152 151 153 154 154 OYKFIUDBHTAGDAACDIXAMRPAGQZTAACVAOWLYX WABBTHJJANEEFAAMEAACBSVSKALLPHANRNPKAZF YRQAQEAXJUDFOIMWZSAUCGVAOABMAYAAZJDAL JACINEVBGAOFHARPVEJCTQZ9PJLEIQWNAHRBUIAS >\ZMWAAAWHACAXHXQAXTDPUTYGSKGRAVLGKIM FUOFAAKYFGTMBLYZIJAAYAUAACXDTVDACJSIUFMO TXWAMQEAKHAOPXZWCAIRBRZNSOQAQLMDGUSGB AKNAAPLPAAAHYOAEKLNVFARJAEHNPWIBAYQRK UPDSHAAQGGHTAMZAQGMTPNURQNXIJEOWYCREJD UOLJCCAKSZAUAFERFAWAFZAWXBAAAVHAMBATAD KVSTVNAPLILAOXYSJUOVYIVPAAPSDNLKQRAAOJLE GAAQYEMPAZNTIBXGAIMRUSAWZAZWXAMXBDXAJZ 79 80 The Measurement of Mental Traits in ECNABAHGDVSVFTCLAYKUKCWAFRWHTQYAFAAAOH GAAQYEMPAZNRTIPXGAIMRUSAWZAZWXAMXBDXAJZ ECXABAHGDVSVFTCLAYKUKCWAFRWHTQYAFAAAOH UOLJCCAKSZAUAFERFAFWAFZAWXBAAAVHAMBATAD KVSTVNAPLILAOYSJOUVYIBPAAPSDNLKRQAAOJLE AKNAAPLPAAAHYOAEKLNVFARJAEMNPWIBAYAQRK UPDSHAAQGGHTAMZAQGMTPNURQXXIJEOWYCREJD TXWAMQEAKHAOPXZWCAIRBBZNSOQAQLMDGUSGB FUOFAAKYFGTMBLYZIJAAVAUAACXDTVDACJSIUFMO SNZMWAAAWHACAXHXQAXTDPUTLGSKGRKVLGKIM JACIXEYGBAOFHARPVEJCTQZAPJLEIQWXAHRBUIAS YRQAQEAXJUDFOIMWZSAUCGVAOABMAYDYAAZJDAL AYKFUIDBHTAGDAACDIXAMRPAGQZTAACVAOWLYX WARBTHJJAEXEEFAAMEAACBSVSKALLPATXRMPKAZF El It is evident that such a test as this can readily be arranged into any number of sets of uniform difficulty. We adhered to the plan previously followed in the use of this test (which was first employed by Bourdon with a different arrangement of test material) of occasionally placing two or three A's in juxtaposition. But we do not believe that is well to use these com- binations, owing to the tendency of some subjects to cross two or three A's with one stroke of the pencil. Very few of our subjects did this, however, as they were especially cautioned to draw a separate line for every A. 2. XATURE AND PURPOSE OF THE TEST This test furnishes a convenient device for measuring the speed and ac- curacy of perception and motor reaction and of the subject's ability to sustain attention. In order to react properly to the test the subject must be able to discriminate between the different letters, and particularly to recognize all of the successive A's, and he must be able rapidly to draw a line through the A's as soon as they are perceived. He must make a series of discriminative reactions. To continue these reactions for any length of time involves a con- siderable strain on the attention. The eyes must be focused continuously on the lines; no line must be missed; the letters must be accurately recognized, particularly the A's, which is rendered somewhat difficult because of the lack of spacing between the letters. If the attention relaxes some A's may be overlooked or wrong letters may be crossed. Moreover, the subject must, keep the problem continuously in mind. The test can, therefore, we believe, be legitimately classified as a test of perceptual discrimination, sustained at- tention, and selective motor reaction. Normal and Epileptic School Children 81 One of the evident advantages of this test is that it can be used with individuals widely differing in mental capacity. It can be given to young or low grade children who are merely able to recognize the printed capitals, par- ticularly the A's, and who are able to hold a pencil and to draw a line, while adolescents and probably adults are not able to make perfect scores under the prescribed conditions. 3. DIRECTION'S FOR GIVING THE TEST General Directions: See pages 15 to 17. Order of Giving Tests: As before. Time: 100 seconds precisely (one minute and 40 seconds). Place papers upside down on the pupils' desks, explaining that they must not be turned over before the signal is given. Instructions to Pupils: "On the reverse side of the paper you will find all the letters of the alphabet printed in capitals. The letters do not follow in natural order; they come without any definite order at all. Now, the moment I say 'turn' you are to turn the papers rapidly and draw a line through every capital A; you must draw a separate line for each A. You must take the lines in order. Begin with the first one, on the left end, then go to the second line, left end, etc. Do not miss any A's or any lines, or mark any other letters, as that will count against you. Mark all the A's you can until I say 'stop.' Let me show you what I mean." (Demonstrate what is wanted by placing a row of figures on the board and marking out the 3's. Thus : 6429856^54^45109873561642^456^.) ■Now. ready!" "Turn!" "Now. stop!" "Turn you papers. Sign your names." Ranking: Record the number of A's crossed. Record the number of A's missed. Record the number of wrong letters crossed. Count up to the last A crossed. How to grade the papers in per cents on a combined qualitative-quanti- tative basis: For every A crossed off credit .5% (one-half per cent). For every A missed deduct .5% (one-half per cent). For every wrong letter crossed deduct .5% (one-half per cent). M T 6 82 The Measurement of Mental Traits in 4. TABLES TABLE 19 Speed and Accuracy of Perceptual Discrimination: El— 5. Normal Children No. Ae. Sx 1 % 2 % 1 J 1 3 Vi J 2 4 Vr J 8 5 Vc J 4 % 1 % J 1 -2 j i % Gain Ave. % 2 7 R 14. o qfi 0 00 U 1 97 J. • C 1 01 7 1 90 1 . ZU 22. 1 1 oo 1 .uu is ^ lo.D AO .00 10. z 20 2 1 OA 1 Ofi lfi Q lo .y D .0 2 7 fZ VJ 23.5 OA R Z4 .D 1 39 1 . oc 26.5 1 oo 1 . uu 25 .2 QR .yo ZO .D 1 10 1. JZ Ot; o ZD . U 26 8 i no 1 1. ZD zd .y 1 4 7 A IT AV. 18.9 20. 1 ~\ 9R 24 - 1 1 10 1 1U 23 .6 C7 OO c ci . D yo 9 Ofi Z.UO 23.5 1.16 1 OA OO 0 ZZ Z 1 . 1 g Q 0 D 17 t; 1/0 00 7 ZU. / 1 99 1 . CC 00 0 1 OQ i uy 21 .9 Q3 .yo 93 7 ZO I 1 Vi 1 . 10 1Q 1 iy . i 22 6 1.16 1 At oo o ZZ .U 1 .D q Q 0 c 23 3 01 fi Zl .0 96 24.6 1 \A 1 . 14 32.0 1 .01 97 n ci U .yo 99 ^ ZZ . D 29.4 1.34 1 1£ 1 . 10 ot; 7 ZD . / .y 11 Q 0 A \t iy.u 01 Q _ i .y 1 . 10 00 Q 1 11 1.11 24 .9 1 i\a 1 .U4 9>< A Z4 .0 1 Ofi 1 uo 90 1 ZU. 1 24.5 1.22 1 0.1 1.01 00 o Zo .U 1 .4 Q 0 Q y R o jy.o Ifi 7 JO / i ni OA C 1 9^1 I . Z4 23.3 QO .yo OK 3 ZD .0 1 11 1 11 90 fi ZU . 0 24 8 1.21 1 Ofi oo o Zo .U 2 A 4 Q y f* Kj 18.8 24 . 1 1 9^ 1 ZD oc n ZO U 1 1Q i iy 23 . 6 OO ot; C ZD .0 1 07 1 .U/ 91 0 Zl .0 24.5 1.13 1 00 1 .ot. oo o Zo U 1 Q i . y 7 Q y A \r 19.0 21 . 2 1 ■ ID 25 . 2 1 22 23.5 QO .yz or q ZD 0 1 .09 01 O Zl .u 24 6 1.16 1 1A 1 ,oA oo o Zo.U 1 Q i y 3 10 B 27.5 24.2 92 25 8 .99 25.8 .99 25.5 .99 26.2 25.7 .97 93 26 0 5 q in 1U OC £ ZOO 09 R OZ .0 1 91 OC Q 00 .0 1 1 o 1 . 1U 37 .8 1 OR 1 ud Ofi c; 00 D .yo 9Q fi zy . o 37 1 1.26 1 Ofi 1 OO 00 Q oo . y 9 J Z . 4 c D 10 _U A \T 27 . 1 OQ 0 zy . z 1 HQ i .uy Ofi Q OU.O 1 . U4 31 .8 1 oo 1 . UZ 0.1 n 0 1 .u .yo 97 Q ci .y 31.4 1.11 1 1 R 1 . 10 OQ Q zy .y 1 O 1 U q a 11 R 30 . 1 1/1 o o4 U 1 1 9 1 . 1 Z 29.0 fl7 .07 39 3 1 0^ 1 . OO 07 9 o / . z .y4 79 1 OZ 1 38.2 1.17 1 01 1 .LI Oil O o4 . U 1 Q 1 .0 Q 11 c 35 . 1 1 n q 4U .0 1 17 1.1/ 41 .3 1 Ol 1 Ul 40.6 yo 40 . 0 1 17 1 . 1 ( 07 O o/ • y 42.9 1.13 1 OA 1.04 jIO 7 4U. / O fi O . o 11 11 Av. 33.7 38 9 1 15 37.9 .97 40 5 1.08 43.6 1.10 36 3 41.6 1.14 1.30 38.9 3.1 5 12 B 31 4 34.0 1 08 36.0 1.10 43 5 1 17 39.7 .95 32 6 40.5 1 15 1.28 36 3 2.2 3 12 34 .8 39 . 8 i it; 1 ID 44 .3 1.11 41.7 .97 ^i t; 41 .D QO yy 37 0 Of .0 41.5 1 11 1 1 Q i iy 40 .4 1 fi 1.0 8 12 A 17 rt V . 32-7 nc o 00 Z i in 1 . Iv 07 Q 0/ .0 1 10 1 1U 42 7 1 07 1 -U/ ^o 0 4U .0 yo 04 . 4 40.9 1 14 1 9t; 1 .ZD 07 fi 0/ .0 1 Q i .y 4 13 B 30- 7 33.8 1 21 40 6 1 16 45.0 1.12 43.6 I 95 33.9 44.3 1.31 1.42 41.9 3.4 2 13 G 32 0 37 7 1 18 44 7 1 19 42 c .93 45 7 1 09 34.9 44.1 1.26 1 43 40 5 3 4 6 13 Av. 30.6 35 4 1.19 42 6 1.17 44.1 1.06 44.0 .99 34 2 44 2 1 29 1.42 41.4 3.4 4 14 B 30.0 35 6 1 20 34. 7 .97 36.8 1.04 40.2 1 10 32.8 37.2 112 1.31 34.9 2 4 7 14 G 39 3 45 2 1 14 46.4 1.04 48 9 1.10 53.2 1.02 42.3 45 4 1.14 1.37 47.5 3.4 11 14 Av 35 0 41 7 1 17 41.2 1.01 45 3 1.08 49.5 1.05 38.8 42.4 1.20 1 35 42.9 3.0 2 15 B 51.0 58.5 1.14 58 5 11 64.7 1.04 69.2 1.06 54 7 66 9 1 22 135 60.9 5 2 2 15 G 41.2 52 0 118 47.5 1.05 50 5 1.06 57.5 1.13 44 3 53.9 1 25 1.39 49.1 4.6 4 15 Av 46.1 56.3 1 15 51.2 1.07 57.6 1.05 63.4 1.09 49.5 60.4 1 22 1.37 55.0 4 9 1 16 B 51.5 52.5 1 01 54 0 1.02 57.5 1.06 52.0 55.7 1.07 1.11 53.8 2.0 3 16 G 44 2 50.5 1 16 52.3 1.12 56.8 1 06 63.5 1.12 47 9 60.2 1 25 1.39 54 4 . 5.5 4 16 Av 46.6 51.1 1.09 52.3 1.12 56.1 1.05 62.0 1.10 48- 9 59 0 1.24 1.32 54.3 4.6 1 17 B 41 0 47.0 1 1 14 51.0 1.08 53.0 1.03 60.5 1.14 44 0 56.7 1 28 1.47 50.5 4.8 2 17 G 61.0 64.0 1.08 67 2 1 05 72.2 1 07 74.2 1.02 62.8 73.2 1.14 1.21 68.4 3 8 3 17 Av 54.3 55.5 1.11 61.8 1 06 65.8 1.06 69.6 1 06 1 57.2 67 7 1.19 1.38 62.4 4.1 Ave. B 27.6 30 8 1.14 30 8 1.08 35.2 1 05 35.5 1.03 1 29 1 35.4 1.17 1.29 31.9 2.1 Ave. G. 33.9 38.6 1.16 41 3 1.08 42.2 1 03 •J5 2 1.06 ! 36.1 43 7 1.19 1.32 39.7 3.0 Ave. B & G 30.8 34 8 1.15 36 2 1 08 39.0 1.04 40 5 1.05 32.7 39 8 1.18 1.31 36.3 2.5 Normal and Epileptic School Children 83 TABLE 20 Speed and Accuracy of Perceptual Discrimination: El— 5 Epileptics Su. Ae. Sx. 81 10 B 102 10 G 104 10 G Av G 99 11 G 79 12 B 84 12 B 91 12 B Av. B 108 12 G 93 13 B 86 13 B Av. B 97 13 G 77 14 B 89 14 B Av. B 80 15 B 82 15 B 83 15 B Av. B 103 15 G 78 16 B 88 16 B 85 16 B Av. B 9fi 16 G 98 16 G 106 16 G 100 16 G Av. G 87 17 B 90 17 B Av. B 95 17 G 105 18 G 107 21 G 101 24 G Ave. B . Ave. G. Ave B&G 16.0 20.5 26.0 23.2 24.0 23.5 22.0 23. 5 23.0 0. 23 5 11.0 17.2 19.5 28.0 25.0 26.5 29.5 14.0 16.0 19.8 19.0 14.5 17 5 22.0 18.0 0. 29.5 12.5 12.5 13.6 5.5 0. 2.7 0. 37.0 22.5 0 18.2 15.9 17.1 % X '2 Q % f\ 4 % ft - % f fi 1-2 % 4-5 % 17.0 1 06 24 5 1.44 16 0 20 7 320 1 56 25 5 79 32 '5 1 27 36.0 1 10 26 2 34 2 33 0 1 26 34 .0 L03 51 5 1 51 44*0 85 29 5 47 7 32 5 1 .41 29.7 91 42 .0 1.39 40.0 .97 27 8 40 9 33 5 1 39 36 '5 1 08 48 0 1.31 28 7 42 2 31 0 1 31 31 .0 1 00 33 5 1 08 15 0 .44 27 2 24 2 23*0 1 04 29 0 1.26 30 5 1 05 34.0 1 11 22.5 32 2 23 0 .97 31 0 1 34 38 5 1 24 39 0 1 01 13 2 38.7 25 6 1.10 30.3 1.20 34 1 1 12 29*3 85 20.9 31 .7 45 18 5 4 11 16 0 86 14 5 90 2.2 15.2 21 0 '89 28.5 1 35 23 5 82 29 0 1.23 22.2 26.2 15 * 5 1 40 10 0 64 10*5 1 05 7.5 71 13 2 9 0 18 2 1.14 19 2 99 17 0 .93 18 2 .97 17 7 17.6 22 0 1 12 23 .5 1 07 21 5 91 20.7 22 5 34 .0 L21 34 5 1 01 35.0 1 01 43 5 1.24 31 0 39 2 29 5 1 18 39.5 1 33 28 0 70 32 0 1 14 22 2 30.0 3j 7 1.19 37 '0 1.17 3T5 85 37 7 1. 19 26.6 34.6 35 0 1 18 34.0 '97 30'o .88 32 5 1 08 32 2 31.2 28.0 2.00 20 '5 73 30 5 1 48 25 0 81 21 0 27.7 25.5 1 59 30 5 1 19 34 0 1 11 24.0 70 20 7 29 0 29 5 1 59 28.3 96 31.5 1 15 27 1 86 24 6 29.3 8.5 44 24 5 2 88 22 5 91 22.5 1 00 13 7 12.5 22 o 1 51 21 0 95 27 0 1 28 92 5 83 18 2 24 7 42 0 2". 40 4L5 .98 4L5 1.00 21.5 .51 29.7 31.5 21.5 .97 30.5 1.37 27.0 .88 28.0 1.03 21.7 27.5 28.5 1 .62 31 .0 1 10 31.8 1 05 24.0 79 23.2 27.9 11 .5 18 0 1 56 0 00 27 0 5 7 13 5 25.5 .86 43 5 1.70 47.0 i'08 27.5 45.2 12.0 96 19.0 1 .Do 20.5 1 fl7 1 .U/ £.6.3 1 no 10 O 21.0 1.68 20.5 .97 22.0 1.07 20.5 .93 16.7 21.2 14.8 1.32 20.7 1.24 21.5 .96 29.2 1.03 14 1 25.3 18.0 3.27 18.5 1.02 22.5 1.21 21.5 .95 11.7 22 0 0. 0. 0. 0. 0 0 9.0 3^27 9.2 i'62 11.2 1 21 10.7 .95 5 8 11 0 10.5 10. .95 12 0 1.20 0. 11 7.5 .20 41 0 546 23.0 .56 53 0 2.30 22 2 38 0 22.0 .97 25.0 1.12 27.0 1.08 30.5 1.12 22 2 28 7 15.0 10.5 .70 19.0 1 80 8 5 44 7.5 13 7 24.6 1.49 26 6 1.04 26 8 1.05 24. 9 .94 21.3 25 8 18.5 1.06 23 7 1.78 23.9 1 06 29 1 1.09 15 1 26.7 21.9 1 33 25.2 1.42 25.5 1.06 26.9 1.01 19.4 26.2 1 29 1 30 1.61 1.45 1.67 88 1.43 2 93 1 74 6 90 1.18 .68 .93 1.03 1.26 1 35 1 30 96 1 31 1.40 1 22 91 1 35 1.06 1.26 1.22 2.37 1.64 1 76 1.26 1.75 1.88 188 i 1 71 1.29 1.82 1.34 1.94 1.62 1.53 1.75 1 69 1.72 2.00 .63 1.54 1.65 1.27 123 68 97 1 10 1.55 1 28 1 41 1.10 1.78 1.50 1.46 1.17 1.55 1 22 1.27 1 34 1.59 1 80 1.64 1.67 3.90 3.90 1 43 1.35 1.49 1 55 1.51 5. RESULTS (See Tables 18 and 19) COMPARATIVE EFFICIENCY The average efficiency for all the normals in all the five sittings (column 1-5) was 36.3% and for the epileptics 23.3% . In this test the epileptics proved to be 64% as capable as the normal children. The average for the epileptics was a trifle better than the average for the youngest normals, the seven-year-olds, and 37% as good as the average for the oldest group of normals. The boys' capacity, based on the general average at the foot of the col- umn for the five sittings (1-5), was appreciably inferior to the girls among the normals, but slightly superior to the girls among the epileptics. The normal boys did 80% as well as the normal girls, while the epileptic girls did 92% as well as the epileptic boys. Based on the sex averages for each sit- 84 The Measurement of Mental Traits in ting (at bottom of columns), the normal girls did better than the normal boys in every sitting, while the epileptic girls did poorer than the epileptic boys in four sittings and better in one. Based on the sex averages in the different ages for sittings 1 to 5 (next to the last column in the tables), the normal girls did better than the normal boys in eight ages, and poorer in two, while the scores were the same in one age. On the other hand, the epileptic girls did poorer in four ages and better in two than the epileptic boys, while no comparison can be made in the other ages. The ability measured by this test improves decidedly from age to age. Based on the averages for the two sexes for sittings 1 to 5 (next to the last column in the tables), there is an improvement in seven ascending ages and a loss in two, while equal averages occur once. The losses occur in ages 12 and 16, in which 62% and 75%, respectively, of the children are classed as dull and average. The equal scores are in ages 8 and 9, in which, re- spectively, 72% and 57% of the children are classed as dull and average. The difference between the youngest and oldest groups of subjects amounts to 40.2% (62.4%-22.2%). In this test the highest score does not occur, as before, in age 15, but in age 17. Here again the epileptics fail to show consistent improvement with in- creasing chronological age. This test decidedly came within the range of capacity of the epileptics. Only one epileptic failed to make any score in all the sittings, while one failed in two sittings and three in one sitting. One "bright" child in the fifth grade, curiously, made so many errors that his score was -7.5% in one sitting and -2.5% in another. His record was eliminated from the final tabulation as an abnormal record. None of the other normal children failed to make a score in any sitting. Both the normals and the epileptics made their highest score in the fifth or last sitting and their lowest score in the first sitting, and both groups im- proved consistently in each successive sitting. The average monthly gain made by each subject during the five sittings (last column) amounted, in absolute units, to 2.5% both for the normals and for the epileptics. The relative amount of improvement, i. e., relative to the size of the average scores — was greater for the epileptics than for the normals, as shown by the following percentages of improvement in the second sitting compared with the first, in the third compared with the second, etc. (based on the indices of improvement given at the bottom of the tables) : comparative improvement M 2 4 Normals . Epileptics 15% 33% 8% 42% 4% 6% 5% 1% 31% 51% Normal and Epileptic School Children 85 Relatively the epileptics gained more than the normals in all the sittings except one, where they gained only 4% less. In most of the sittings the difference is very considerable, amounting to 18% in the second sitting com- pared with the first (in favor of the epileptics), 34% in the third compared with the second, 2% in the fourth compared with the third, 20' I in the fifth compared with the first, and 44% in the average of the fourth and fifth compared with the average of the first and second. In the aggregate the relative gains are considerably larger for the epileptics. The indices show a gain in all sittings for both the epileptics and normals. The relative improvement declined in each successive sitting with one excep- tion each for the epileptics and normals. Based on the average monthly gain for sittings 1 to 5, one epileptic made no improvement and two showed losses, which is 10% of all the epi- leptics, while three normals, or 4% of all the normals, showed a loss. The absolute monthly gain for the five sittings was one and a half times larger for the normal girls than for the normal boys (3.% vs. 2.1%) and two times larger for the epileptic girls than for the epileptic boys (3.5% vs. 1.7%). Based on the average indices of improvement at the foot of the columns the normal girls gained more than the normal boys in two sittings (or four, including / {:f and/ ?), less than the boys in one sitting, while the gains are equal in one sitting. The epileptic girls gained more than the epi- leptic boys in all the sittings (including / }:f and f \) except one, where the boys gained more. The gain in the fifth sitting over the first sitting was 3% more for the normal girls than for the normal boys (32% vs. 29%), and 6% more for the epileptic girls than for the epileptic boys (55% vs. 49%). The gain based on the average of the last two sittings compared with the average of the first two (/ }:■§) was 2% more for the normal girls than for the normal boys (19% vs. 17%), while for the epileptics it was 60% more for the girls than for the boys (94% vs. 34%). In this test, therefore, the tendency of the girls was to gain more than the boys. When each age is compared with the next succeeding age there is an increase among the normal pupils in the amount of improvement in five ages and a decrease in five ages also, based on the average monthly gain in the final column. Based on the index /flf, there is an increase in five ascending ages and a decrease in four, while the gains are equal in two ages. The results for the epileptics vary, entirely as before, with the individual. 6. CONCLUSIONS 1. Our normal children in this perception-reaction-attention test did only about 80% as well as the pupils in our dental experiment. This superiority of the dental group is not wholly due to the absence of children 86 The Measurement of Mental Traits in under ten in this group, for the superiority exists in five of the seven ages which can be directly compared. This has proved to be a very valuable test for the comparative measure- ment of the strength of a fairly complex trait (involving perceptual, motor and attentive factors) under easily controlled and standardized conditions, and for the measurement of the amount of improvement which may occur in the trait with time, in individuals six or seven years old and over. 2. The epileptics were about two-thirds as efficient as the normal chil- dren in the A-test. 3. The normal girls were superior to the normal boys in the A-test, not only in this experiment but also in the experiment on the dental group. But the epileptic boys were slightly superior to the epileptic girls. 4. The ability to cross out A's tends to increase from age to age among our normal pupils. The exceptions which occur may be due to a preponder- ance of backward children in certain ages or to the fewness of the subjects in the different ages. The achievement of our oldest group is almost three times as great as the achievement of our youngest group. 5. That the individual differences in crossing A's are very considerable is apparent from the following extreme variations between the lowest and highest individual scores in each sitting: 1 2 3 4 5 For the epileptics. 0 to 37.0% 0 to 42.0% 0 to 41.5% 0 to 51.5% 0 to 53.0% For the normals 10% to 63.0% 15% to 65.0% 15 to 70.5<* 13.5% to 73% 14% to 76% The individual differences in each age, based on the average score for all sittings (1-5), among the normals are: Age 7 8 9 .10 11 12 13 14 15 16 17 Difference 11.6 22.2 15.0 11.4 28.5 16.9 26.1 27.1 10.9 10.3 20.1 6. In this test the average monthly improvement was greater for the epileptics than for the normal pupils. While the absolute gain was the same in both groups, the relative gain was greater for the epileptics in all the sit- tings except one. Since this is not a complicated test, we should suppose that the normal pupils would not need to spend much time learning how better to do the test, so that they ought to be able to function near their maximum at the outset. A part of the epileptics' gain, on the other hand, probably is due to learning better how to perform the test — how rapidly to look for the A's and to make rapid and accurate strokes. Both the normal pupils and the epileptics made consistent, though diminishing, gains in this Normal and Epileptic School Children 87 test from sitting to sitting. Only 10% of the epileptics and 4% of the normals failed to show a monthly improvement. 7. Both the normal and the epileptic girls improved more from month to month in the A-test, both absolutely and relatively, than the boys. In our earlier experiment the girls, again, improved slightly more than the boys. 8. There is little, if any, correlation between the amount of monthly improvement in this test and the age of the normal child. It would seem that the older pupils should not improve as much as the younger pupils, as they ought to be able to function nearer their maximum at the beginning. 9. In spite of its simplicity, our age and improvement curves both emphasize the value of the A-test for measuring mental capacity (specifically a certain type of complex trait) and mental improvement, in individuals of a very wide range of ability. Chapman's correlation between initial and final efficiency was greater for his two cancellation tests (.75 and .85) than for the opposites test, but less than for his three other tests. VII. OBSERVATION Range of Visual Apprehension: Fl to F5 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS Ten "objects" were placed on each of the five white cardboards, measur- ing 22 x 28 inches. Each cardboard was surrounded on the reverse side by a wooden frame so as to prevent buckling. The cardboard was supported by the right hand from a cross piece placed horizontally across the middle of the reverse side. A piece of cloth, somewhat longer and wider than the card- board, was fastened to the top of the frame work. A small wooden rod, about a foot longer than the width of the cardboard, was fastened to the bottom of the cloth. When the cardboard was held before the subjects the cloth shielded all the objects. The exposure of the objects was effected by the experimenter grasping the rod with the left hand and rapidly throwing the screen upward to the reverse side of the cardboard. When the time for the exposure had expired the shield was again jerked rapidly forward over the front of the cardboard. Considerable preliminary practice was engaged in in making exposures, in order to increase the speed of the movement as well as to make the exposure time uniform. It is evident that with the method employed it would be impossible to make the exposures absolutely uniform in length, but the variations were inconsiderable, except in the case of one sec- tion each in Fl (normals), F3 (epileptics), and F5 (epileptic girls), where the exposures were slightly lengthened because the cloth caught momentarily when it was being flung back to its original position. The cardboards contained the following objects:1 'We give these descriptions since these cardboards have not been included in the set prepared by C. H. Stoelting Co. 88 90 The Measurement of Mental Traits IN The Measurement of Mental Traits in Normal and Epileptic School Children 93 j OBJECT CARD BOARD NO 5 | Fl: a tree nine inches tall, with green branches, planted in a white tub; a post card with a child in a red suit; a brown bottle, 1% x 5 inches; the letter C, 3 inches high; a hatchet with a red blade and black handle, 10y2 inches long; a reddish metal rooster, 3 inches high; a spool of black thread, 1% inches long; a red disc, 4 inches in diameter; the figure 5, 3 inches high; a shovel, with a black blade, 3 inches long, and a red handle, 8 inches long. 94 The Measurement of Mental Traits in F2: a red top, with green rings, 2V£ x 3% inches; a woman in wood, with red body and white and bluish head,- 2% x 6% inches; the figure 6, 2% inches high; a brown pencil, 6% inches long; a post card with the word Yale on a blue ground; a black disc, 4 inches in diameter; a black cap, 3 x 4 inches (including the brim in the latter measurement) ; the capital H, 2 x 3^ inches high; a yellowish cob pipe, 2x/± x6 inches; and a tin dust pan with a picture and colored effects in red, blue and yellow, 6^4 x 7V4 inches (includ- ing the handle). F3: the figure 2, 3 inches high; a basket made of green fibre with red edges and bands, 4 inches in diameter; a black comb 1% x 7 inches; a post card with a child and red and greenish embellishments; a screw driver with a black handle, 6 inches long; the capital E, 3 inches high; a box of safety matches, 1% x 1% inches; a yellow-bluish ball with black band, IV2 inches in diameter; a light green disc, 4 inches in diameter; and a silk U. S. flag, 4% x 7 inches. F4: a black metal horse with red band and golden reins, 3% x 5 inches; a blue disc, 4 inches in diameter; a red stocking, IV2 x 5Vi inches; the figure 9, 3V2 inches long; a red rose, 3% inches in diameter; the letter A, 3 inches high; a green metal chair, inclined forward, 3x5^ inches; a horn handled knife, 6 inches long: a post card with the face of a laughing boy, 0V2 x 6V2 inches; and a black brush, 1 x 7% inches long. F5: a doll with a red suit and yellowish hair, 2x7 inches; the letter D, 2% inches high; a post card with a house in black and white. 3x5 inches; a fork with a black handle, 7% inches long; a saw with a black handle; a nickeled bell with a black handle, 21/^x4V2 inches; a yellow disc, 4 inches in diameter; a wooden cow in white and black, 4x6 inches: a gilded watch and chain, the watch being 2% inches in diameter; and the figure 8. 2% inches high. It will be observed that each cardboard contains one illustrated post card, one figure, one capital letter, one colored disc (including the black disc), and six solid objects. It was thought that the chances of securing cardboards of equal difficulty would be furthered by this systematic arrangement. But. of course, it would require an extended research to determine whether or not the cardboards are absolutely uniform in difficulty. No objects were selected which were thought not to come within the daily experience of the pupils tested. In order that inequalities might not arise through unequal difficulty in the spelling of the names of the objects in the different sets, the pupils were instructed to make a rapid crude sketch of the objects whose names they were unable to spell. The epileptics and the younger pupils oc- casionally made such sketches. It was not difficult in the vast majority of the cases to determine from the study of the complete record what the draw- Normal and Epileptic School Children 95 ings were intended to represent. It was explained that it was only necessary to write the names of the objects, and not a description of each object. Many pupils, however, sometimes wrote brief descriptions, such as a red rooster, a red hatchet, or a spool of black thread. In a few instances the records were slightly inadequate. For example: the colored disc was men- tioned usually as a ring, or ball, or spot, but no reference was made to the color. The colored discs were usually referred to as the red (or green, etc.) spot, or red ring, or red circle. The post cards were frequently mentioned without any reference being made as to what was on the cards. Such blanket responses might deserve full credit during the first sitting or two, but hardly thereafter. Again, instead of mentioning the card, merely the picture on the card was referred to, as a girl, a house, or a laughing face. Sometimes only a part of the object was mentioned (e. g., a pot, for the tree in the tub), or one aspect of an object (e. g., brown, for the brown bottle), or a quality was recorded which was found in several objects (e. g., red in Fl, which could equally well refer to the red suit worn by the girl on the post card, or the red blade of the hatchet, or the reddish tint on the rooster, or the red disc). In the latter case it was often possible by a process of exclusion to determine just what the recorded quality referred to. Sometimes the objects were mis- apprehended or misnamed (e. g., the cap was called a hat, the hatchet a hammer). When the misnaming did not seem to be very significant in view of the brevity of the exposure, we felt justified in giving full credit (e. g., for " string" referring to thread or a spool of thread, or "music box" referring to the top). Sometimes the less important part of an object was mentioned (e. g., pot for the tree in Fl). It is evident, in view of the many sources of error, that it is no easy matter to make up a list of suitable materials for a test with a very limited time exposure, and that it is not always easy to eval- uate the papers. We found it necessary to allow only partial credit for a considerable number of responses. The credits which were given for unusual responses are given in the third section. All of these objects could be easily seen from any seats in the room which were occupied by the pupils. All the pupils said they had no difficulty in seeing the objects, although we suspected that three or four suffered from imperfect vision to such an extent that they could not clearly see all the objects. The records were made on blank slips of paper, 4% x 7 inches. 2. NATURE AND PURPOSE OF THE TEST This crude tachistoscopic test furnishes a measure of the rapidity or range of visual observation or apprehension during a limited exposure. It must not be forgotten, however, that this trait was not measured in the present test by means of oral responses, but by means of written words or 96 The Measurement of Mental Traits in graphic representations. Some of the pupils, particularly the younger normals and the lower grade epileptics, might have done better if the re- sponses had been oral instead of graphic. Ample time was allowed, however, in which to write the replies. Many of the pupils did not need more than half of the time allowed for the responses. One of the obvious advantages of this test is its wide range of applica- tion, even when the responses are written. The capacity to observe or appre- hend develops very early in the life of the child. Any child is able to take the test who can write the names of the objects or make a crude sketch of them. Tests similar to this have frequently been given in the schools with a view to increasing the range of apprehension. 3. DIRECTIONS FOR GIVING THE TEST Time: Two seconds for the ready signal; five seconds for the exposure; 60 seconds for writing the replies. Instructions to Pupils: "Beneath this cloth cover you will find a num- ber of things fastened to a white cardboard. I am going to uncover them by lifting the cloth rapidly and throwing it behind this frame. The moment I uncover the objects you must look at them just as carefully and rapidly as possible, for I will only let you look at them a short while. Then when I cover the objects again by jerking the cloth back, you must write down everything that you can remember seeing. You do not need to describe the things that you see, but only write the names of them. If you cannot spell the words or name some of the objects, you can make a simple little picture or a drawing, instead of writing the name. But you must not spend too much time on these drawings. Just make them enough like the thing so that I will know what you mean by them. Now, remember that you must look quickly so you'll see all of the things, and then try to write down everything you saw if you can. But don't write until I say so." "Now, ready!" (Expose objects and after five seconds say) "Write." "Now, stop!" "Turn your papers, sign your names." Ranking: Give 10% credit for every object correctly named. Deduc- tions may be made for the enumeration of wrong objects, but this was not done in this experiment. It was thought that incorrect responses would be sufficiently penalized by the withholding of all credit. It was, however, necessary to give partial credits for answers which were not wholly wrong, or which were only partially satisfactory. As we have already stated, some answers mentioned only one aspect of an object. By a process of exclusion based on the answers given correctly, it was usually possible to determine what objects were referred to by the partly correct answers. But this was Normal and Epileptic School Children 97 not always the case. The amount of credit which we have allowed for these incomplete answers is confessedly more or less arbitrary. We have in a few cases, taking the brevity of the exposures into account, allowed full credit for responses which were not entirely adequate. It is evident that the proper scoring of the papers was frequently difficult, because of the inadequacy of the replies. It is possible that the replies would have been more adequate had we emphasized more fully that the answers must indicate clearly just what was seen in everything that was on the card. But it is also evident that had we told the subjects that they must state what they might see on the post card or that they must mention the letter or the figure which they might see on the cardboard, we would have suggested the presence of a letter, a figure or a post card on the cardboard. Fortunately some of the partial and unusual responses were only given by one pupil. The following are the partial credits, or perfect credits for rather inade- quate responses, which we have allowed in the different sets: Fl: post card, without mentioning what was on the card, 10%. The same credit was given in F2. but only half credit in the later sets for the same response. Hammer (for hatchet). 8';. Red (for disc), 10%. Brown (for bottle). 3^c. Picture (for post card). lO'/J . Letter (without mentioning C), 3%. Black (for spool, hatchet for shovel), 3'/t . String or thread (for spco'.). 10';. Pot or flower (for tree), 5%. Whistle (for rooster), 10%. Fl. reproduced one month later: ring or ball (for red disc), 5%. Doll (for picture on post card). 10',; • Card, 10%. Number or letter, without mentioning C or 5, 5%. Green (for tree), 5%. F2: Doll, or little man. or baby, 10';. Black (for the cap), 3%. Black ball (for black disc), 10',' . Circle, 3%. Rattle (for top), 5%. Hat vfor cap). 10%. Stick (for pencil), 5%. Flag (for Yale pennant on post card), 107c; Yale (for post card), 10%. Shovel (for dust pan), 10%. Figure (for 6), 3%. Bell (for top), 5%. Picture (for card), 10%. Toy (for top), 3%. Music box (for top), 10%. Whistle, 3%. Red (a number of objects were red), 3%. F3: Green (for basket or card), 5%. Circle, 3%. Box (for match box). 10' . Picture, 10%. Button hook, or stick of wood (for screw driver), 3' ; . Chisel, or awl, or tack puller (for screw driver), 10%. Stilleto or dagger (for screw driver), .5%. Girl (on post card), 10%. Post card, 5%. Cloth (for flag), 3%. Top (for ball), 3%. Number (for 2), 3%. Paper cutter (for comb or screw driver), 5%. Green spot or green cloth (for disc), 10%. Red (for either flag or card), 3%.- F3. reproduced one month later: Red (for post card or flag), 3%. Picture, 10%. Circle (for green disc), 3%. Card (for post card), 5%. Number or letter (without mentioning 2 or E), 3%. M T 7 98 The Measurement of Mental Traits in F4: Circle (for blue disc), 3%. Blue (for blue disc), 10%. Girl (for boy), 3%. Leg (for stocking), 10%. Arm (for stocking), 5%. Red (for red nose), 3%. Hat (for part of picture on card), 5%. Pink (for stocking). 3%. Black (for brush), 3%. F5: Red spot or red (for doll), 5%. Round color, or color, 5%. Orange (for disc), 10. Ball (for disc), 3%. Post card, 3% (less credit was given in this sitting for "post card," since the subjects would probably take it for granted that a post card would again be shown). Horse, or donkey, or dog (for cow), 5%. Man (for doll), 5%. The following data may be separately recorded: 1. Number of things named correctly (meaning by "things" everything exclusive of the post cards, figures, letters, and colored disc). 2. Whether the post card was given (if so indicate by a plus sign). 3. Whether the letter was given (indicate by a plus or minus sign). 4. Whether the number was reproduced (indicate by plus or minus). 5. Number of objects or items mentioned which were not on the cards. 6. Whether the colored disc was mentioned (indicate by plus or minus sign). Normal and Epileptic School Children 99 4. TABLES TABLE 21 Range of Visual Apprehension: Fl— 5 Normal Chldren No. Ae. Sx k 1 f\ 4 % 5 % f\ 1- 7 2 4- ot / 5 m f\ 1-5 % Ga in Ave. Vc 2 7 Boy 20 0 11 5 57 19 0 1 65 15 0 79 32 5 1 50 15 7 23 7 i 18 1 12 19 6 3.1 2 7 Girl 45 0 30 0 75 42 5 1 05 45 0 1 19 39 0 90 41 8 42 0 98 85 42 1 —1.5 4 7 Av. 32 5 17 6 66 30 7 1 25 30 0 1 05 36 1 1 10 29 7 32 8 i 05 94 30 8 .8 8 8 Boy 31 2 30 3 98 32 2 1 09 41 8 1 29 38 3 1 05 30 8 38 7 i 36 1 22 34 2 1.8 3 8 Girl 21 0 27 6 1 46 42 0 1 73 29 3 95 35 3 1 36 21 3 33 0 i 30 1 81 31 0 4.4 11 8 Av. 28 4 29 6 1 12 34 9 1 27 38 1 1 19 37 5 1 13 29 0 37 1 i 34 1 38 33 3 2 5 3 9 Boy 30 0 45 0 1 62 39 3 94 50 0 1 43 46 6 90 33 0 48 3 i 47 1 72 41 0 4.5 4 9 Girl 31 2 31 0 1 03 39 3 99 40 7 1 56 34 2 93 29 7 37 5 i 50 1 48 34 0 .8 7 9 Av. 30 2 36 6 1 27 39 3 97 44 7 1 50 39 5 91 31 2 42 1 i 49 1 58 37 0 2.5 4 10 Boy- 32 5 31 6 1 22 46 0 1 27 50 7 1 23 51 2 1 03 32 8 51 0 i 64 1 84 42 5 5 0 3 10 Girl 36 6 41 o 1 14 45 0 1 14 53 3 1 18 39 0 72 38 8 46 1 i 18 1 06 43 0 5 7 10 Av 34 2 36 3 1 18 45 5 1 20 51 8 1 21 46 0 90 35 4 48 9 i 44 1 51 42 7 3 0 3 11 Boy 40 0 36 6 1 16 33 3 94 50 0 1 58 54 3 1 09 38 3 52 1 i 42 1 72 42 8 3 5 8 11 Girl 38 5 55 3 1 60 51 3 95 61 0 1 19 41 5 79 46 9 54 0 i 15 1 40 50 7 3 0 11 11 Av. 38 9 50 0 1 48 46 4 94 58 0 1 30 49 5 88 45 5 53 5 i 23 1 49 48 5 3 1 5 12 Boy 40 0 47 2 1 24 47 6 1 06 47 0 89 47 8 1 19 43 6 45 2 i 04 1 26 45 2 1.9 3 12 Girl 34 3 48 6 1 46 54 3 1 47 45 0 92 61 0 1 56 41 5 53 0 i 38 1 84 48 6 7.5 8 12 Av 37 8 47 7 1 32 50 1 1 21 46 1 90 52 7 1 33 42 8 48 1 i 17 1 48 46 8 4.0 13 Boy 53 2 43 3 1 00 48 2 1 03 45 7 95 50 7 1 10 C D A Q o 95 97 49 5 c . O 2 13 Girl 40 0 50 0 1 25 57 5 1 12 50 0 90 58 5 1 22 45 0 54 2 i 21 1 46 51 2 4 6 6 13 Av. 48 8 46 6 1 10 51 3 1 06 47 1 93 53 5 1 14 48 4 50 2 i 04 1 13 49 9 2 0 4 14 Boy 54 5 65 7 1 23 58 7 92 68 3 1 23 56 6 78 60 1 62 5 i 04 1 05 60 5 8 7 14 Girl 45 4 52 2 1 26 47 6 1 05 52 2 99 50 8 1 06 48 1 51 5 i 08 1 15 49 9 1.7 11 14 Av. 48 7 57 2 1 25 52 5 99 57 1 1 06 52 6 97 52 5 55 5 i 06 1 11 53 7 1.4 2 15 Boy 70 0 70 0 1 00 70 0 1 00 8i l 0 1 14 70 0 87 72 5 75 0 i 57 99 66 0 0. 2 15 Girl 32 5 62 5 2 36 33 0 53 52 5 1 58 6(1 0 1 13 47 5 56 2 i 19 2 30 48 1 3.1 4 15 Av 51 2 65 0 1 90 54 0 69 66 2 1 36 65 0 1 00 50 0 65 6 i .38 1 65 57 0 1 5 1 16 Boy 60 0 40 0 66 60 0 1 50 63 0 1 05 50 0 61 5 i 23 1 05 55 7 1.0 3 16 Girl 60 0 56 5 1 03 56 o' i 00 58 3 1 06 64 0 1 09 56 8 61 1 i 02 1 09 59 1 1,0 4 16 Av 60 0 51 0 90 56 0 i 00 58 7 1 11 63 7 1 08 55 1 61 2 i 07 1 08 58 3 1.0 1 17 Boy 60 0 60 0 1 oo 60 0 i no 70 0 1 16 63 0 90 60 0 66 5 i 10 1 05 62 6 .7 2 17 Girl 65 0 63 0 78 69 0 l 22 75 0 1 09 76 5 1 03 64 0 75 7 i 17 1 19 69 8 — 1. 3 17 Av. 63 3 61 5 89 66 0 i 15 73 3 1 11 72 0 92 62 0 72 6 i 15 1 14 67 4 - .3 Ave . B 41 1 41 3 1 11 43 6 l 06 49 7 1 20 51 6 1 04 41 2 50 7 i 26 1 30 45 5 2 2 Ave .G. 40 3 39 5 1 36 48 9 l 10 51 7 1 14 49 7 1 04 39 9 50 7 i 19 1 39 46 1 2 3 Ave . B&G 40 7 40 4 1 25 46 3 i 08 50 8 1 17 50 6 1 04 40 5 50 7 i 23 1 35 45 8 2.3 100 The Measurement of Mental Trait '8 IN TABLE 22 Range of Visual Apprehension: Fl— 5 Epileptics /Ye. C v OX . l % / 1 3 % £ 3 4 % /il 5 % I - •k 2 4 - 3! p . Ave 1 ~ 5 % 81 10 Boy 20 0 ■ * ■ ■ • 40.0 2 AA 00 40 0 1 . oo 20 0 40 0 n 2 An 1 10 2 Ail Oil 33 3 10 0 102 10 Girl 40 0 30 0 .75 30 0 .66 40.0 2 00 50 0 1 25 35 0 45 0 1 28 1 25 36 0 2.5 104 10 Girl 25 0 30 0 1.20 23 0 1.10 43.0 1 30 30 0 69 27 5 36 5 1 32 1 20 32 2 1 2 Av. Girl 32 5 3^ 0 .97 26 5 .88 ^ j 1 65 40 0 97 31 2 40 7 1 30 1 22 34 1 1.8 99 11 Girl 15 0 40 0 2.66 30 0 75 53 0 1 76 27 5 41 5 1 50 3 53 34 5 12.6 79 12 Girl 15 0 20 0 1.33 30 0 1.50 10 o 33 0 ■ 00 17 5 5 0 28 00 15 0 -3.7 84 12 Boy 50 0 43 0 86 30 0 .69 00 5 2 00 47 0 78 46 5 53 5 1 15 94 44 0 - .7 91 12 Boy 55 0 60 0 1.09 40 0 66 70 o 1 75 4'i 0 57 57 5 55 0 95 72 53 0 -3.7 Av Boy 40 0 41 0 1.09 33 3 .91 46 g 1 36 29 0 45 40 5 37 8 79 55 37 3 -2.7 108 12 Girl 20 0 2U 0 1.00 20 0 1.00 3 o p "»E £ . ~ o 5>ca»*o „,TJ l) re fi"S 9 £73 2 -2— ~ > P 01 =_ a a _.p a „.= .kSVh Z ■| °- S « <= « 2 5 2.c_ o; u p £ Z c m E — -° 5 p 3 oj « g — c ££$"ESi E >.Z-< — £0=.&o,02 J-a P o/-= Jr.— DOO 3 t£ O — - _ . - _ " • . ■ coo ■ cm ■ *—> co ■ -cm ■HNinfonNmw c— - to in ■ in cm cm co cm mm cm co cm co • CM CM ■ IT) CM -co CM — ( Z CM CM CM ■COCMCMCMCOCMCOCOCOCMCOCMin CO CD CM in t- -torn CO COCO inmincMcocMcocMcom -cmcoc— r-totocoininin in in c— com mco incMCM'-ocotOimnr- cococ—coaocoin r— cm CM~CMCMCMCOCOCOCM^COCOCMCOCOCMCOCM^CM^C^ •m cm .-«to cm coco io co co co co m in co coco co m m m cmco into coco -mco coin- m t— — • co cm to to to co coco'-'Tj'CMcocMm co cm .—• mco ifimuimicwntcic co co co co c— (C»ifliomMC M" cot— r- CM CM CM cm -o ■ co ^ co co co co cm to co in co co t— in in t*» • - i— lOCMCOCO CM CM m co co t- co cm co m • m c f-t -Oifin*f co «— o co co m cm m m in in ■ to o co co co to m co r— co co t— co in r- in in r- cO'HC»;-Hir:ioocoto<£> ■ COO m m m cm Tf co mr- com -tocoinco r- mto m cotom cd)CMCMCMCMCOCOCOCOCMCMCOCMCOCOCO'« co to r- CM CM CM ceo >mo >mo >oqo >cdo >oao >ao >mo >ajo >cqo >coo 5 <<<<<<<<<<< NNt^OOQOCOOlO)OiOOO-4^'-CJNNmmPjTj'*^inmiOtOlOtDr*NN >,T" u) w 1/1 mom 4) 0J 0J > > > «c« 102 The Measurement of Mental Traits in CO « (fl •lf> CO CMTfCMCO^OCOCOCO cm eMCMto^co^TrcococMLncoin^^^cococMCM^rcoeMCMCMCM eri—ie*i m cocm cm — i -CO NMN oo tn to «o m to cm n cm co cm cm cm cm co — • .— < — ■ — co co cm co co cm cm co ~ - -- - : r - i ~ i r i — ■ — in co CM CM CM to to w co • ■ co co co "<*■ cm co to co cm < co cm m cm co co co co - co m m CO '—• * ^4 CO CO CO CO ' CO •—' C CM ^? CM CO CM CM CO <0 CO CM CM ■ CM — ' CO CM CM CM — « CM ~* -*T CO CM ■ CM Tf CM CM CM CM CM -CO CO CO CO CM CM CM >,T" rrr >< >> >>*r >* > >, >,t" "rrrr >. >> >vt* rrr o . -.-.«.= o o o o - oooooooooo c- o o o o - o o o - 'O OOO CM CM CM > CM CO CO CO > LO LO lO £ lO y2 'X> ^^O'-C'^'Xi >r-t^ > f~ 00 — oooo CT»r~ooo> oo>oocn mom ii v tt > > > Normal and Epileptic School Children 103 5. RESULTS (See Tables 20, 21, 22a, and 23) IMMEDIATE RECALL COMPARATIVE EFFICIENCY The average efficiency for all the normals in all the five sittings (next to the last columns in the tables) was 45.8% and for the epileptics 30.9%. The normals were able to give not quite one-half of the objects and the epi- • leptics not quite a third. The epileptics' comparative efficiency amounted to 67' ! of the normals. The average for the epileptics was the same as for the youngest normals, the seven-year-olds, and only 46% as good as for the oldest group of normals. The boys' capacity in this test, based on the general average for all five sittings (given at the bottom of the next to the last column), amounted to 98' ; of the girls' capacity among the normals, while the girls' capacity among the epileptics amounted to only 73% of the boys'. Based on the sex averages for each sitting (at the bottom of the columns), the normal boys did better than the normal girls in three sittings and poorer in two sittings, while the epileptic girls did poorer in all of the five sittings. Based on the sex aver- ages in the different ages for sittings 1 to 5, the normal girls surpassed the normal boys in seven ages, while the boys surpassed the girls in four ages. W ith the epileptics the boys excelled in five ages and the girls in one, no com- parison being possible in the remaining ages. The power of observation, as measured by this test, improves uniformly from age to age. Based on the averages for the two sexes for sittings 1 to 5, there is an improvement among the normal children in all the ages except one, namely age 12, in which there is a slight loss. In this age, however, only 62', of the pupils were classed as average and dull, against 72% in age 11. The difference between the averages for the youngest (who made the lowest score) and oldest subjects (who made the highest score) amounts to 36.6% (67.4%-30.8% ) . There is no consistent gain with increasing chronological age among the epileptics. Xot a single epileptic failed to score in all the sittings, while only one failed in two sittings (in each case she had a convulsion before the test was given) and four failed in one sitting. One of the latter was incapacitated by a convulsion a short time before the test was given, and one was below normal on this day. One of the normal pupils failed to make a score in all five sittings (No. 5, the "dull" seven-year-old boy in the second grade), and one failed in one sitting (an "average" 8-year-old boy in the second grade). The normals and the epileptics made their highest averages in the fourth sitting, while the normals made their lowest average in the second sitting and the epileptics in the first. 104 The Measurement of Mental Traits in COMPARATIVE IMPROVEMENT The average monthly gain made by each subject during the five sittings (last column in the tables )amounted, in absolute units, to 2.3% for the normals and 1.9% for the epileptics. The epileptics improved 82% as much as the normals did. Based, however, on the index of improvement, which gives the amount of improvement relative to the size of the average score, the epileptics gained more on their own records than did the normals on their records. This is seen from the following percentages of improvement of the second sitting compared with the first, of the third compared with the sec- ond, etc. (based on the indices at the bottom of the table) : Normals 25% 8% 17% 4% 35% 23% Epileptics 43% -3% 19% 4% 43% 25% The difference in favor of the epileptics amounts to 18% in the second sitting compared with the first, 2% in the fourth compared with the third, 8% in the fifth compared with the first, and 2% in the average of the fourth and fifth compared with the average of the first and second. On the other hand, the normals gained 11% more than the epileptics in the third sitting compared with the second, while the relative improvement is the same in the fourth sitting compared with the third. The only loss, according to the indices of improvement, was made by the epileptics in the third sitting compared with the second. If we turn to the figures at the foot of the columns in the tables containing the percentage scores, both the epileptics and the normals lost in two of the successive sit- tings, while they gained in two. (Strangely, the epileptics lost in both of the sittings in which the exposures were made somewhat too long in one section.) The discrepancy between the gain according to the index and loss according to the score in the second sitting for the normals, is accounted for by the fact that the total in column 1 is divided by more than the totals in columns 2 and / 'i, because of the absence of six subjects in the second sitting. Based on the results in the column containing the average monthly gain for sittings -1 to 5, one epileptic made no improvement while ten showed losses, which is 36% of all the epileptics. On the other hand, nine normals made no gains while nineteen lost, which is the same percentage for the normals. The absolute monthly gain for the five sittings was about the same for the normal boys as for the normal girls (2.2% vs. 2.3%), but almost one and two-thirds times larger for the normal boys than for the normal girls (2.3% vs. 1.5%). Based on the average indices of improvement at the foot of the in- dex columns, the normal girls gained more than the normal boys in the second and third sittings, and in the fifth compared with the first, the boys gained more in the fourth sitting and in while the improvement was equal in the Normal and Epileptic School Children 105 fifth sitting compared with the fourth. The epileptic girls gained more than the epileptic boys in the fifth sitting and in the boys gained more in the second, fourth and fifth compared with the first, while the losses are the same in the third sitting. The gain in the fifth sitting over the first was 9% more for the normal girls than for the normal boys (39% vs. 30%), and 12% more for the epileptic boys than for the epileptic girls (49% vs. 37%). The gain based on the average of the last two sittings compared with the average of the first two (f f:f ) was 7% more for the normal boys than for the normal girls (26%) vs. 19%), but 1% more for the epileptic girls than for the epi- leptic boys (26% vs. 25%). When the gain in each age is compared with the gain in the next follow- ing age, the amount of the improvement, based on the average monthly gain in the last column, among the normal pupils increases in five ages, decreases in four, while there is one instance in which the gains are equal. Based on the index / there is an increase in five ascending ages and a decrease in five. The corresponding results for the epileptics are very irregular. ANALYSIS OF DATA In Tables 22a and 23 we have made an analysis of the replies in each sit- ting, in order to determine which types of presentation (the letters, figures, post cards, colored discs, or objects, which we refer to as "things") were the most easily noticed or recalled. This tabulation, unfortunately, is not abso- lutely accurate. Thus we restricted the use of the category "things" to everything on the cardboards except the post card, number, letter and colored disc. One of the two teachers, however, who helped us to analyze the data classified the post cards as things. She did not analyze many of the papers. It was impossible to correct this mistake, because we did not learn of it until the original data were inaccessible (they were left at the New Jersey State Village for Epileptics). Again, we are not absolutely certain that this assist- ant properly recorded the answers relating to the post cards, letters or num- bers. These were to be scored as plus (or as "mentioned") only provided the illustration was mentioned on the card, and only provided the particular letters and numbers shown were correctly designated. It is possible that the assistant recorded plus for general answers like "post card," "letter," and "number." Owing to the inaccessibility of the original data and the inability to reach the assistant in question we cannot definitely settle this point. Moreover, it has frequently been very difficult to rule that some things which were mentioned were not on the cardboard, because the "things" mentioned could sometimes be regarded as aspects of some of the objects which were actually on the cardboards. In presenting the data we desire to call attention to these unremovable defects. We shall refer merely to the gross results. 106 The Measurement of Mental Traits in The average number of "things" correctly mentioned was as follows for each of the normal children in the five successive groups: 2.7, 2.9, 2.7, 3.1 and 3.2. For the epileptics the corresponding figures are: 2.3, 2.5, 2.3, 2.2 and 2.4. It will be noticed that the normal children were able to name, on the whole, only about one-half of the '"things" (of which there were six on each card), while the epileptics named less than one-half. The average number of mentions made of the post cards by each normal pupil is as follows in each of the five successive sittings: .6, .4, .4. .7 and .6. The corresponding figures for the epileptics are: .2. .2. .3. .5 and .5. Taken all in all, about one-half of the normal pupils mentioned the post card and only about one-third of the epileptics. The epileptics, however, did almost as well as the normals in the last two sittings. Full credit was given for some responses which might be considered slightly inadequate, such as "picture," flag (F2), man's hat (F4), and book with girl (F3). The average number of mentions per normal pupil of the letters is as follows in the five successive sittings: .2, .2, .3. .3 and .4. The corresponding figures for the epileptics are .03, .07, .03, .07 and .03. Less than one-third of the normal pupils, except in one sitting, mentioned the letters, while the number of epileptics who mentioned them is negligible. The average number of times that each normal pupil mentioned the numbers is as follows in each of the five successive sittings: .2, .3. .6, .5 and .3. For the epileptics the figures are .0, .07, .2, .2 and .10. On the average, a little over one-third of the normal pupils mentioned the numbers, while only one-fifth of the epileptics mentioned the numbers in the two sittings in which they made the best record. The average number of mentions per normal pupil of the colored discs is as follows for the five successive sittings: .3, .3, .4, .3 and .2. The cor- responding figures for the epileptics are .06, .03, .0, .1 and .03. In general, less than one-third of the normal pupils mentioned the colored disc and only a negligible number of the epileptics. It should be mentioned that in tabulat- ing the results for the colored disc no credit was given a considerable number of indefinite or incomplete responses which, no doubt, frequently referred to the colored discs. Thus no credit was given the following responses among the normals. In F2: circles (6 times); black objects (twice). In F3: ring (once). In F4: circle (once) and red circle (once). In F5: ball (three times); circle (3 times), red circle or red sphere (twice each). Among the epileptics the following were not credited in F5: ball (2 times); O with no color indicated (twice). On the other hand, the following indefinite responses from the normals were credited as correct: In F2, black (3 times). In F3, green (6 times, although there were other greenish objects on the cardboard). In F4, blue (7 times). In F5, yellow (once). For the epileptics the following was given credit: in Fo, orange. Normal and Epileptic School Children 107 Comparing the above results, we may conclude that "things" were the most readily recalled by both groups (in harmony with an earlier finding in which the memory of things was compared with the memory of words), while the post cards were almost as frequently recalled by the normal pupils. The letters, numbers and colored discs were recalled about equally easily, but they were distinctly more difficult of recall than the things or post cards, more par- ticularly so for the epileptics. The letters and numbers probably attracted less attention than the objects, because they were inherently less attractive and smaller. The colored discs, however, were as large as many of the "things" and some of the colors were quite bright. The average number of objects mentioned by each normal pupil which were not on the cardboards is as follows for each of the five successive sit- tings: .2, .1, .3, .3 and .2. For the epileptics the figures are: .1, .0, .1, .1 and .1. The number of wrong objects mentioned averages less than one- third of an object for each normal child, and only about one- tenth of one object for each of the epileptics. The reason that the epileptics made less mistakes is not due to the fact that they were keener observers, but to the fact that they made fewer responses than the normal children and responded merely to objects as a whole. The number of errors does not seem large, considering the brevity of the exposure. It should be explained, however, that we allowed credit for re- sponses that were not strictly correct. Thus the following responses were considered to be adequate, in the sense that what was mentioned was con- sidered to be on the cardboards: Fl, normals, coat (for the child on the post card), wood (for the hatchet or tub of tree), and pink (for the reddish rooster). F2, normals: orange (for the yellowish pipe, or the yellowish tint on the dust pan), white (for the white face on the woman), brown (for the brownish pencil), green (for the green ring on the top), and purple (for the blue on the dust pan or on the woman). F2, epileptics: wood (for the woman). F3, normals: black (for the black comb), and file and nail file (for the comb). F4, normals: red (three times, for the red stockings), white (twice, for the white post card), green (for the green chair), wire (for the wire chair), black (for the horse), yellow (for the golden reins). F5, normals: baby (twice), boy (three times, for the doll), wood (for the wooden cow), white and black (for the white and black cow), horse (10 times), donkey (five times), and dog (three times — for the cow), brass or iron (for one of the metallic objects), yellow (for the yellowish hair on the doll or the watch and chain), eyes, or nose, or mouth or feet (for parts of the doll), tail (for the tail of the cow), and gilt (for the gilded watch and chain). F5, epileptics: horse (five times, for the cow), and boy (twice, for the doll). Among the objects which were incorrectly given are the following: Fl, normals: watch, "wash," flag, "gut," cake, box, blue, and eight other ob- 108 The Measurement of Mental Traits in jects the names of which cannot be given because of the inaccessibility of the original data. Fl, epileptics: dog, "catch," and one other object. F2, normals: horse, wax, brush, pit, apron, pot, "fore board," coal, C, and D. F3, normals: top (twice), bottle, brush, hatchet, ribbon, thread, spool of thread, rattle, pencil, nail file, file, pencil, knife (four times), C, "Ch," "red eb" and seven other objects. F3, epileptics: knife, man, "three," 4 and tree. F4, normals: ball (six times), scissors (twice), oval, dog, B, E, M, 1, 3, bath tub, green sphere, pen, ball, ring, cup, bug, orange, "fall," "a bug," and seven others. F4, epileptics: washstand. red ball, dog, lady and girl. F5, normals: knife (four times), man (twice), red ring (twice), scissors (twice), ball, glass, hammer, brush, "forty," scenery, chair, hatchet, top. E, and two other objects. F5, epileptics: ribbon, ball, hut, cat, clown, knife, scissors, and "frick ball." DEFERRED RECALL Retention of Objects after an Interval of One Month Without warning, the subjects were asked in the February sitting before F2 was given to write down all the objects which they could recall from Fl, which was shown one month earlier. In April they were similarly asked to make out a written list of the objects shown in March. The time allowed for writing was the same as in the original tests, 60 seconds. Normal and Epileptic School Children 109 TABLE 24 Range of Apprehension after the Lapse of One-Month: January i Fl ) in February; March 1 1Z 4 9 4 . £ l 3 1 . 0 86 13 B 0 0. 0. 10. 0. nn .UU 0 . ft u n u . n u ■ Av. B .... 4. 2. .50 5. 2.5 2. 9 9 Z -Z OO 9 1 Q O. 97 13 G 33. 5. . 15 10 2.00 "_ 20.0 16 0 ic n io .U Q3 • yo . nu 17 O -A Q 4.0 77 14 B 70. 70. 1 .00 70 1.00 24. .34 69 0 - 1 .87 70.0 oy . o . y* .ys fiU 7 oy . / ~ 3 89 14 B 25 5 .20 40. 8.00 0. nn .UU 55. 15 a 7 c 4/ D q i k O.IO 2 20 31 2 10 • ■ • • Av. B 47 5 37.5 .60 55. 4.50 12. . 17 62. 2 .87 4Z D DO .0 9 nfi Z UO 1 • i . oy 50 4 4.0 80 15 B 45.0 45.0 1 .00 44.0 .97 0. nn UU 50 45 .0 a7 n 4 i . U 1 f\A 1 U4 1 11 46 0 1 A i . 0 82 15 B 65.0 50.0 .76 75 0 1 .50 19. 25 65 0 3.42 57 5 70.0 1 21 1 .00 63.7 0. 83 15 B 30. 0 35.0 1.16 30 0 .85 0. .00 45 0 32 5 51 D i i r 1 ID 1.50 35 0 c • ■ ■ ■ Av. B 46.6 43 3 97 49.6 1 . 10 6.3 no uo 53 3 1 AO O .4Z 45 0 013 1 . 15 1 20 48 2 9 9 103 15 G 5.0 0 00 0 0. 0 2.5 ft u. nn .UU 'oo 1 2 -1 fi 1 . 0 78 16 B 60.0 73.0 1 21 49 0 .67 10. .20 60.0 6.00 66 5 54 5 81 1 00 605 0. 88 16 B 45 0 72.0 1.60 75.0 1 04 10.0 .13 68.0 6 80 58 5 71 5 1.22 1.51 65.0' 7 6 85 16 B 70 0 39.0 .55 65.0 1 .66 15.0 .23 65 0 4 .33 34 D 65.0 1 1Q i . iy 09 • yz 59 7 -l fi 1 0 Av H 58.3 61 3 1 12 63.0 1.12 11.6 . 18 64 3 5.71 57 . 5 Do -O 1 .U/ i i x 61 7 o . it . 96 16 G 5. 0. .00 0. 0 0 2 5 0. .00 .00 l!2 -1.6 98 16 G 29.0 5.0 \i7 5. 1.00 0 .00 29.0 2 5 08 00 11.3 -14 5 106 16 G 0. 10. 10. 1 00 0. .00 10. 5.0 10 0 2.00 7.5 3.3 100 16 G 39.0 40. 1.02 55 0 1.37 9. .16 60. 6\66 39 5 57 5 1.45 i .53 48.5 7. Av. 18.2 16.6 51 17 5 83 3.5 .38 17.5 3.33 17 5 17 5 .88 .48 17.5 -.27 87 17 B 0 0. 4. 0. .00 0 0. 2. 1.0 0 90 17 B 5. 0. HO 0 0. 0. 2 5 0. 00 !oo 1.2 -1.6 Av. B 2.5 2 1.2 1 1.1 -.8 95 17 G 0. 0. 0. 0. 0. 0. 0. ' ^50 0. 0. 105 18 G 10.0 0. "oo 0. 0. 5. 5 2.5 50 3.7 -1.6 107 21 G 0. 28.0 10. .35 5. 50 10. 200 14 10 .71 1.00 12. 3.3 101 24 G 5. 15. 3.00 0. .OC 0. 0. 10 2.5 25 6.2 0. Ave B 32.6 33 4 88 39.4 1.6? 8.0 .14 39.0 3 51 33.0 39.1 1.28 1.46 36.2 2.2 Ave G 17.0 11.3 .68 11.0 1 04 1 5 .27 14 9 2 88 14.2 13.0 .76 .66 13.6 -.9 Ave B & A 25.3 19 6 80 26 4 1.2] 5.2 .18 27.7 3 36 24 2 26.7 107 1.11 25.5 .7 TABLE 27 Percentage of Normal Pupils who Correctly Reproduced the Illogical Pairs of Words Normal Children o c 1 c .= o 1—1 No. 2 T m No. 17 O 1 No. 18 2 P Dollar— than o W T >. s s No. 2 o No. 17 S JL No.18 2 ? Rain— and o V > id jt 1 t/i 3 £ re No 2 (M 1 No. 17 U. 1 No.18 re T s o B No.20 5 o c 1 u 93 X. CO w No. 2 1 CO No. 17 2 1 05 No.18 E I •a c No.20 Ave B Ave G Ave B & G .3 .3 3 .7 .5 .6 .6 .6 6 .02 .02 .02 .3 .2 .3 .8 7 .7 4 4 .4 06 .05 05 .3 .5 .4 .6 .6 .6 .5 .6 .6 .2 .3 .2 .5 .4 .5 .6 .8 7 .2 .6 .4 .1 .1 .1 118 The Measurement of Mental, Traits in TABLE 28 Epileptics o c 1 c J= o —1 No 2 ^* 1 in No. 17 Q 1 U No. 18 c CO 1 o Q No.20 z ? May— is to o 7 No 17 X No.18 T3 c a 1 e CO OS No.20 z P James— have CM 1 No. 17 1 No.18 Z ? Snow— that z P Esther— now ; to | 1 CO No.17 Z 1 SO No.18 s ■o B No.20 Ave. B . . 1 .1 3 0 2 .1 .17 .0 .1 .0 .2 0 Ave. G .07 07 1 0 0 07 0 .0 .0 j ■* .07 .0 :! 0 f 0 Ave.B & G .1 1 .2 .0 1 .1 1 .1 .06 .06 .0 .1 .1 03 .0 5. RESULTS (See Tables 25 to 28) IMMEDIATE reproduction comparative efficiency The average efficiency of the normal group for all the sittings, except the fourth, which contains the illogical set, was 68.1% and for the epileptic group 25.5' i . The normals recalled almost fourteen associates while the epileptics averaged only a little over five. The epileptics proved to be only 37% as efficient in this test as the normals. The average for the epileptics was only 60' , as high as the average for the youngest normals, the seven-year-olds, and only 29' < as high as> the average for the oldest group of normals, the 17-year-olds. It is not possible to say whether this large difference was due solely to the inability of the epileptics to retain the associations. It is probable that it was at least partly due to their difficulty in writing the words. The average of the normal pupils for the set of illogical associates, G4, is only 52' ; as high as their average for the four sets of logical associates, while the epileptics' average for the illogical set is only 20% as high as their average for the four logical sets. It must be remembered that each of the four logical sets contains four pairs of illogical associations. In Tables 27 and 28 we have tabulated the results separately for these pairs. We have, however, given only the general averages for all the boys and girls and the combined averages for both the boys and girls for sittings 1, 2, 3, and 5. The complete data are accessible for any one wishing to use them for each epileptic and for each chronological age for the normal pupils. The data for A-L and 14-G in G4 were incom- plete for the epileptics and are therefore not given. The figures may be read as the average number of correct mentions per pupil for each combination, or as the per cent of pupils who correctly reproduced the combinations. The Normal and Epileptic School Children 119 For For For 5—4, E— D, Dollar — than, 11—10, I— H, Rain — and, 7—2, K— F, Snow — that, and and and 3—7. B— N. Wind — must. .71 .50 .12 .11 .15 .00 average number of correct mentions per pupil were as follows (a perfect score would, of course, have been 1.) : For John — not, Mary — is, James — have, and Esther — now. Normals 42 Epileptics 11 The normals did somewhat better on the numbers (71%) which were given out of order (5 — 4, etc.) than they did on all the pairs in the four sit- tings (68.1%)) but they did decidedly poorer on the three other illogical pairs. It is noteworthy that they did decidedly poorer on the pair of illog- ical word associates which came at the end of each series (dollar — than), than on the corresponding illogical pair which came at the beginning of the series (John — not). The same is true for the epileptics. The difference may be due to the factor of primacy. The epileptics did relatively much poorer than the normals, both in the set containing nothing but the illogical associates, and in the illogical pairs which occur in the four logical sets. It has been said that mental defectives can memorize illogical facts about as well as logically coherent facts, but this does not prove to be the case in this test with our epileptics. The capacity of the normal boys in this test, based on the averages of sittings 1. 2. 3 and 5, amounted to 96% of the capacity of the normal girls. On the other hand, epileptic girls did only 37.5% as well as the epileptic boys. The comparative efficiency in the illogical set (G4) was, again, slightly higher for the normal girls than for the normal boys (who were 97% as efficient as the girls), while the epileptic girls were decidedly inferior to the epileptic boys. In fact, only three epileptic girls made any score at all. The comparative records of the boys and girls on the illogical pairs in Gl, G2, G3 and G5 do not differ materially among the normal children, while the epileptic boys are decidedly superior to the epileptic girls, as shown in Tables 27 and 28. Based on the sex averages for each sitting, given at the foot of the per cent columns, the normal girls did better in four sittings and poorer in one than the normal boys, the differences being small in most sittings, while the epileptic girls did decidedly poorer than the epileptic boys in all the sittings. Based on the sex averages for the different ages, for sittings 1, 2, 3, and 5 (third column from the right side), the normal girls excel in seven ages and the normal boys in four ages. Among the epileptics the boys excelled in four ages and the girls in only two, while no comparative data are available in the remaining ages. 120 The Measurement of Mental Traits in The capacity to retain associations under the conditions of this test, im- proves from age to age among normal children. Based on the averages for the two sexes for sittings 1, 2, 3 and 5, there was an improvement in seven ages, and a loss in three, namely ages 9, 12 and 17. In age 9, 57% of the pupils were classed as dull and average as against 72% in age 8; in age 12, 62$ were classed as dull and average as against 72% in age 11, while in age 17 all were classed as dull and average. It is evident that the exceptions are not satisfactorily explained by a preponderance of average and dull pupils in the ages concerned except in the case of age 17. The difference between the averages for the youngest pupils (who also made the lowest score) and the oldest pupils amounts to 42.7% (85.7%-43.0% ). The highest average was made by the sixteen-year-olds. There is no consistent gain among the epileptics. One epileptic did not score a single point in all the sittings, five failed entirely in four sittings, three in three sittings, three in two sittings, and seven in one sitting, five of the latter being in G4. One of the normal pupils (No. 5) failed completely in all five sittings, while two "dull" pupils in the second grade failed in one sitting (G4). Excluding G4, the normals gained in all three of the successive sittings, while the epileptics gained in two of the three, based on the average scores at the bottom of the per cent columns. Both the epileptics and the normals made the highest score in the fifth sitting, while the normals made the lowest score (exclusive of G4) in the first sitting and the epileptics in the second. comparative improvement The average monthly gain made by each subject during the five sittings (last column in the tables) amounted, in absolute units, to 2.6% for the normals and .5 for the epileptics. The improvement made by the epileptics amounted to only 19% of the improvement made by the normals. Based on Gl, 2, 3 and 5, the normals improved 3.9 % and the epileptics .7% , or only 18^5 as much as the normals. Based on the index of improvement, which expresses the improvement as a fractional part of the average scores, the normals gained more in three indices than the epileptics, and lost less in one, while the epileptics gained more in two sittings than the normals, as shown by the following per cents of improvement of the second sitting compared with the first, of the third compared with the second, etc. 2 3 4 5 5 n-s 1 2 it 4 1 1-2 Normals 5% 14% -48% 131% 20% 15% Epileptics -20% 21% -82% 236% 11% 7% The difference in favor of the normals amounts to 25% in the second index (gain in the second sitting over the first), 34% in the fourth, 9% in Normal and Epileptic School Children 121 the fifth sitting compared with the first, and 8% in the average of the third and fifth sittings compared with the average of the first and second. On the other hand, the epileptics gained 7% more in the third sitting compared with the second, and 105' r more in the fifth sitting compared with the fourth. According both to the indices of improvement and the scores in the per cent columns the normals lost only in the fourth sitting, while the epileptics lost in the second and fourth. Based on the results in the column containing the average monthly gain for sittings 1, 2, 3 and 5, seven epileptics made no improvement, while eleven lost, or 60% of all the epileptics, while five normals made no improvement and three lost, which is 23% of all the normals. The absolute monthly gain, based on the average gain for sittings 1, 2, 3 and 5, was about the same for the normal boys as for the normal girls (3.9% vs. 4.0%), but it was two and a third times greater for the epileptic boys than for the epileptic girls (a monthly gain of 2.2% vs. a loss of -.9%). Based on the average index of improvement at the foot of the columns, the normal boys gained more than the normal girls in \t f, \ , f and ?:|, while the girls gained more than the boys in :] . The epileptic boys likewise gained more than the epileptic girls in f (smaller loss), |, ], f and while the girls lost less than the boys in ^. The gain in the fifth sitting over the first was 3% greater for the normal boys than for the normal girls (21 % vs. 18%), and 80% more for the epileptic boys than for the epileptic girls (a gain of 46% vs. a loss of 34%). The gain based on the average of the third and the fifth sittings compared with the average of the first two (/ ?:§) was 9% more for the normal boys than for the normal girls (20% vs. 11%), and 52', more for the epileptic boys than for the epileptic girls (a gain of 28% vs. a loss of 24% ). Comparing the average monthly gains, based on 1, 2, 3 and 5 for each successive age, we find that the amount of the improvement increases among the normal pupils in six ascending ages and decreases in four. Based on the index (/ ?Zf ), there is an increase in four ascending ages and a decrease in six. Among the epileptics the gains or losses vary with the individual irrespective of chronological age. DEFERRED REPRODUCTION Retention of the Sequents or Antecedents After the Lapse of One Month (The completed data are not available for tabulation). In February, April, and May the pupils were asked to reproduce the as- sociations contained in the tests given one month earlier. In Gl and G4 the experimenter read the antecedents and the subjects wrote the sequents, while in G3 the experimenter read the sequents and the subjects wrote the ante- cedents. The subjects had not been informed that these tests of retentiveness 122 The Measurement of Mental Traits in would be given. The average score made by the normals in Gl and G3, re- spectively, was 28.4% and 16.2%. This is 45% and 23%, respectively, of the scores made by the normals in the immediate recall in the same tests one month earlier. The corresponding scores for the epileptics were 13.% and 2.5%, which is 51% and 9.4% of the original scores made one month earlier by the epileptics. The epileptics did not lose quite as much on their previous score in Gl as did the normals, but decidedly more in G3. It is noteworthy that both the normals and epileptics did much worse in G3 than in Gl. We cannot conclude offhand that this was due to the reverse order of recall used in G3, because two sets of word associates had been presented before G3 was given, while no sets had been presented before Gl was given. It is apparent that this circumstance might explain the discrepancy. The difficulty of recall in G3 was increased because of the possibility of confusing the associates in G3 with the associates in Gl and G2. The epileptics' efficiency in the original recall of Gl and G3 amounted to 40% and 38%, respectively, of the normals' efficiency in the immediate recall of the same tests. In the deferred recall the epileptics' efficiency in Gl and G3 amounted, respectively, to 45% and 15% of the normals' efficiency in the deferred recall of the same sets. Comparatively, the epi- leptics did somewhat better than the normals in the deferred recall of Gl, but decidedly worse in G3. The efficiency of the epileptics, based on the average score of Gl and G3 (7.6%), amounted to 34% of the normals' average for the same sittings (22.1%). In the illogical set, G4, not a single epileptic was able to give a single associate, while the average score of the normals was only 1.04' < . which is not quite 3% of the original score made by the normals in this set. 6. CONCLUSIONS 1. The immediate memory span as determined by this test, in which the subjects supply sequents in writing to a list of antecedents read by the experimenter, the antecedents and sequents having first been read together, was only a little over one-third as good for the epileptics as for the normal pupils, the normals reproducing a little more than two-thirds and the epi- leptics only one-fourth of the twenty pairs. 2. The memory of illogical associations was decidedly poorer than the memory of logical associations among both the normal and epileptic pupils, with the possible exception of the illogical numbers for the normal pupils. Contrary to the usual statement that mentally deficient subjects can mem- orize illogical subject matter about as well as logical subject matter, we find that the epileptics did relatively poorer in the illogical series than did the normal pupils, the former doing only one-fifth as well in the illogical list (G4) as in the logical lists of words, while the normals did over one-half as well. Normal and Epileptic School Children 123 3. The ability to reproduce verbal associations in this test was very slightly better for the normal girls than for the normal boys, but decidedly poorer for the epileptic girls than for the epileptic boys. This statement ap- plies both to the logical and illogical associations, although the normal boys did somewhat better, comparatively, in the illogical series. 4. The ability to reproduce the missing sequents of words which were originally presented in pairs increases from age to age among normal chil- dren, the three exceptions to the rule probably being due to the fewness of the subjects in each age. The difference between the youngest and the oldest group is very considerable, the oldest pupils doing about twice as well as the youngest pupils. 5. The individual differences in this test are very considerable. The extreme variations in the different sittings are as follows: 1 2 3 4 5 For the epileptics 0to70% Oto73% 0to75% 0 to 24% 0 to 69% For the normals. 0 to~95% 0 to 99% 0 to 100% 0 to 63% 0 to 100% The individual differences in each age for the normals, based on the averages of the sittings 1, 2. 3 and 5, are as follows: Age 7 8 9 10 11 12 13 14 15 16 17 Difference .... ■■ 66.5 76.5 48.9 29.0 39.5 29.7 14.8 32.9 14.3 8.6 8.0 6. On the whole, the average monthly improvement was greater for the normal than for the epileptic pupils. The absolute gain was considerably greater for the normals; and, while the relative gain in the aggregate was smaller, the normals gained more consistently. The percentage of pupils who failed to improve was larger for the epileptics (60%) than for the normals (23%). 7. There was no significant sex difference in the amount of the average monthly improvement among the normal boys and girls, although the boys gained more than the girls in most sittings. The epileptic boys, on the other hand, clearly improved more than the epileptic girls. 8. There is no correlation between the amount of the monthly improve- ment in this test and the age of the normal children. 9. Both the normals and the epileptics did decidedly poorer in the de- ferred recall (after the lapse of one month) than in the immediate recall, both in the logical and in the illogical scores, but particularly in the illogical series (G4) in which the epileptics absolutely failed and in which the normal pupils almost failed, and also particularly in series G3, in which the ante- cedents were reproduced instead of the sequents. As we have already ex- 124 The Measurement of Mental Traits in plained, the poorer showing in G3 may not be due to the reverse order of re- call, but to the natural tendency, after a considerable lapse of time, of con- fusing the words in G3 with the words previously given in Gl and G2. 10. Our experience with this test, both in group and clinical work, in- dicates that it is one of the most valuable tests in the series. Our group curve for the different ages indicates that it can be successfully employed with normal children between 7 and 17, and probably beyond 17. While it proved too difficult for a number of epileptics, it would undoubtedly prove less diffi- cult as an oral test. For clinical purposes oral responses are probably to be preferred. IX. VISUAL IMAGINATION The Ink-Blot Test: H2 to H5 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS The test materials consisted of the four ink blots reproduced on p. 126, made from zinc block cuts from prints by the C. H. Stoelting Company. Ink-blot No. 6 was used in the second sitting. No. 20 in the third, No. 2 in the fourth, and No. 7 in the fifth sitting. These four blots were selected in the following manner: Seven adults were asked to pick out, independently, the six ink-blots from Whipple's complete set of twenty blots which they consid- ered to be the "best" or the "most suggestive," or the "most fertile in imag- inative possibilities." Blots Nos. 2, 6 and 20 were selected by five subjects and No. 7 by six subjects. The blot which was to be used in the fifth sit- ting received five votes, while the remaining blot (which we had provisionally determined to use in the first sitting) received only two votes, and was there- fore rejected. We do not feel that these blots are equally suggestible or equally rich in imaginative detail and we have accordingly made no attempt to compute indices of improvement from the results for the different blots. 2. NATURE AND PURPOSE OF THE TEST The ink-blot test was suggested in 1895 by Binet and Henri and has since been employed by G. Dearborn, E. Kirkpatrick, Stella E. Sharp and W. H. Pyle, as a test of the fertility of visual imagination, or the degree of suggestibility or imaginativeness. As here administered the test should be classed as a test of active, productive imaginaion, for the subject is asked not merely to look passively at the blots and permit his fancy spontaneously to suggest objects or pictures in them, but he is asked diligently to search out as many resemblances or suggestions as he can find, to construct new forms and to read new interpretations into the patterns. It is, therefore, less a test of reverie than a test of dissociation, abstraction, inventiveness, re- combination or reconstruction (of images). In a word, it is a test — although perhaps a rather crude one — of an important aspect of intelligence. It should be emphasized that in our use of the test the responses consisted of written words. Had the subjects been asked to draw pictures or reconstruc- tions of the blots in accordance with the ideas which they suggested, the results would probably have been different. 125 Normal and Epileptic School Children 127 Distribute the ink-blot cards upside down, requiring the subjects to sit with folded hands, so that they will not turn the cards over before being instructed to do so. Also distribute plain sheets of paper, appropriately numbered (H2, etc.), for the records. Instructions to the Pupils: "On the other side of the small cards you will find a large ink-blot. I want you to look at it, when I tell you to do so, write down on the blank sheet of paper everything that you think you can see in the blot. You should try to imagine that you can see different things or pictures in the blot, just as we often picture things in clouds, or fancy that we see things in the fireplace, although we know that they are not really there. Turn the card around so that you look at it in many positions, for I want you to write just as many things as you possibly can. But do not write on the card with the ink-blot." "Now, ready! Turn the cards over!" "Now, stop. Turn your papers. Sign your names." Scoring: record the number of different "imaginations" or different things written. Our records in this test are purely quantitative. We have merely recorded the number of things suggested by the blots. We have not attempted to evaluate the answers. Needless to say, some of the "imagina- tions" were qualitatively better than others. Many of the imaginations, particularly among the younger pupils, seemed to be quite bootless and far- fetched, and would have been given a low score in a qualitative rating. We have the original records only in H5, which we may analyze qualitatively or otherwise in future. 128 The Measurement of Mental Traits in 4. TABLES TABLE 29 Ink blot Test of Visual Imagination: H2- 5. Normal Children No. Age Sex 2 No. 3 No. 4 No. 5 No. 2-5 Average 9 £ 7 Rrw 5 1 5 1 0 4 0 1 7 9 7 Urn 1 6 0 7 0 7 0 6 0 6.3 A 7 Ave 2 3 4 2 4 2 5 0 4.0 7 t a o Boy 4 1 4 0 4 2 4 1 4.0 Q O a o Girl 5 3 5 0 5 0 2 6 4.5 1(1 a o nvc . 4 2 4 6 • 4 5 3 7 4.1 Q o q Boy 4 5 0 2 6 3 0 2.2 A q un i 8 3 4 3 3 2 2 5 4.3 7 q A \Te> nvc . 6 8 2 2 3 0 2 7 3.4 A J.U Rrw Duy 5 3 5 0 5 0 5 0 4.8 Q O i n C,'\r\ 5 6 5 3 3 6 4 0 4.6 7 in A \r& t\ V C . 5 5 5 2 4 4 4 6 4 7 r> o 1 1 Rrw 2 0 4 0 1 6 2 0 2.4 a o 1 1 V_Tli 1 5 5 5 2 4 0 4 5 4 8 1 1 i l 1 1 1 1 Ave A C O q Q O Q O 3 2 4.3 19 Rrw c o Z 0 4- a o 7 =1 tj 4 6.0 o o 1 9 vjriri < 7 3 6 3 7 6 6 0 6 8 8 12 Ave. 6 0 6 4 8 2 5 6 6^2 4 13 Boy 6 3 4 2 3 2 4 2 4.2 2 13 Girl 4 0 4 0 3 0 3 5 o.o 6 13 Ave. 5 4 4 1 3 1 4 0 4 0 4 14 Boy 6 0 6 2 5 0 7 3 6.4 7 14 Girl 5 1 5 8 3 0 5 0 4.7 11 14 Ave. 5 5 6 0 3 6 5 7 5.4 2 15 Boy 3 0 5 0 3 0 3 0 3.2 2 15 Girl 6 0 10 0 5 0 7 5 71 4 15 Ave. 4 5 8 3 4 0 5 3 5 1 1 16 Boy 3 0 4 0 3 0 3.3 3 16 Girl 8 5 5 6 4 6 7 3 6.2 4 16 Ave. 6 6 5 6 4 5 6 3 4.0 1 17 Boy 4 0 4 0 4 0 1 0 3.2 2 17 Girl 9 0 7 0 4 0 3 0 5.1 3 17 Ave. 6 5 6 0 4 0 2 3 4.5 Ave. Boys. 4 3 4 2 4 1 4 2 4.2 Ave. Girls . 6 0" 5 7 4 2 4 6 5.1 Ave. Boys and Girls 5 2 5 0 4 2 4 4 4.7 Normal and Epileptic School Children 129 TABLE 30 Ink-blot Test of Visual Imagination: H2 — 5 Epileptics Su. Age Sex 2 No. 3 No. 4 No. 5 No. 2-5 Ave. 81 10 Bov 1 4 , 2.5 102 10 Girl 1. 1 . 1.3 104 10 Girl c u . 1 . 4 2. 3.2 Ave . 6. 1. 3. 1.5 2.4 99 11 Girl 9. 6. 7. 7.3 79 12 Boy 0. 0. 2. . 0. .5 84 12 Boy 5. 5. 2. 4 4. 91 12 Rnv 4 4 4 4 4. Ave . 3. 3. 2.6 2.6 2.8 12 Let ull 1 O £ . 2. l 2 1 7 JO Duy 0 £ . 1 4 2 9 9 86 13 Boy 3 4 2. 0. 2.2 Ave 2.5 2.5 3. 1 2.2 97 13 Girl 3 3 ...... 2. 2.6 77 14 Boy 4. 2. 3. 6. 3 7 ftp Rnv DUJ •j o . 1 i i . 1 1 c 1 . o A VP o . o 1 5 2 o o 9 H 1 c. 1 J Rnv 9 2 i 2 1 7 82 15 Boy 2. 3 1 2 2 83 15 Boy 0 . 0 0. 0. 0. Ave 1 .3 1.6 5 1 3 1 2 103 15 Girl 0. 3. 1 , 0 1 78 16 Rnv 1 1 1 2. 1 9 88 16 Rnv Q o . 5 6. Q o ■ d. ^ . j 85 16 Rnv 4 4 1 c; u . O D Ave . 9 fi £ . u fi O . vJ 9 fi o . o 9 96 16 Girl 1, 1. 0. 0. .5 98 16 Girl i. 1. 2. 1.3 106 16 Girl 0. i . 2. 3. 1.5 100 16 Girl 4. 5. 10. 6 6.2 Ave. 1.6 2 3 2 2 7 2.4 87 17 Boy 2. 4 9. 8. 5.7 90 17 Boy 0. 1. 0. o'. !2 Ave. 1 . 2 5 4.5 4. 3. 95 ' 17 Girl 0. 0. 0. 0 0. 105 18 Girl 3. 2. 0. 3. 2. 107 21 Girl 2. 2. 2. 3. 2.2 101 24 Girl ll 5. 2. 1. 2.2 Ave. 1 Boys . . 2 3 2.5 2.3 2 6 2.4 Ave. Girls 2.5 2.5 2.0 2.2 2.3 Ave. Boys and Girls 2.4 2.5 2.2 2.5 2.4 5. RESULTS (See Tables 29 and 30) The average number of suggestions or imaginations, based on the aver- ages for all four sittings given at the bottom of the last column in the tables, was 4.7 for the normals and 2.4 for the epileptics. In other words, the epi- leptics got only 51% as many suggestions out of these blots as did the normal children. The general average for the epileptics was less than the average M T 9 130 The Measurement of Mental Traits in in any age for the normal children, while the general average for the epileptics for each blot was only about one-half as high as the average for the cor- responding blot for the normal children. The epileptics got 60% as many suggestions as the youngest normal group, and 53% as many as the oldest group, based on the averages of sittings 2 to 5. One epileptic failed entirely in all the sittings, three in three sittings, one in two, and three in one sitting, which is 26% of all the epileptics. Only two normal children failed in as many as two sittings, while five failed in one sitting, or 9% of all the normals. We cannot infer offhand that this large difference was due solely to the infertility of the "imagination" of the epileptics. It may at least have been partly due to the inability of the epileptics to write what they imagined, be- cause of the difficulty of writing or of spelling the words. The number of imaginations given by the normal boys was 82% as high as the number given by the normal girls, while the number of imagina- tions given by the epileptic girls was 96% as high as the number given by the epileptic boys. The normal girls got more suggestions from every blot than the normai boys, based on the averages at the bottom of the columns, the greatest difference amounting to 1.7 imaginations (H2), while the smallest difference amounted to only .1 (H4). The epileptic boys got more sugges- tions from two blots and fewer suggestions from one blot than the epileptic girls, while the result is the same for one blot. But the difference among the epileptics does not exceed .4 for any blot. Based on the sex averages for the different ages for all four blots (last column), the girls excelled in eight ages and the boys in three ages among the normal pupils, while among the epi- leptics the girls excelled in two ages and the boys in four. The most suggestive ink blot in the case of the normal pupils was No. 6, shown in the second sitting, and in the case of the epileptics Nos. 20 and 7, which proved to be equally suggestive, shown in the third and fifth sittings. The least suggestive blot for the normals and for the epileptics was No. 2, shown in the fourth sitting. The normals made the highest score in the second and third sittings and the lowest in the fourth and fifth. We may assume that the loss during the last two months represents a difference in the degree of suggestibility of the blots rather than a decline in imaginative fer- tility. At the same time, it is possible that the normal pupils grew more critical as the tests proceeded, and therefore wrote less in sittings 4 and 5. The greatest difference between the averages of the different ink-blots amounts to one suggestion for the normals and only .3% of a suggestion for the epileptics. Based on the age averages for the normal children for all the sittings (last column), there was an increase in the number of suggestions in five ages and a decrease in five ages. The difference between the averages for the lowest and for the highest age amounts to only .5 of a sugf stion. The high- Normal and Epileptic School Children 131 est score was made by the twelve-year-olds and the lowest by the nine-year- olds. Had the records been scored qualitatively the results might have been different. 6. CONCLUSIONS 1. The normal pupils proved to be about twice as suggestible or imag- inative in the ink-blot test as the epileptics. It is possible that the difference would not have been so large had verbal responses been used instead of written responses. 2. The normal girls ranked slightly higher than the normal boys in degree of suggestibility. The epileptic boys were more suggestible than the epileptic girls, but the difference was insignificant. 3. The degree of imaginative suggestibility was probably not the same for the different ink-blots. 4. We find no significant age difference in imaginativeness or sug- gestibility as determined by this test. Kirkpatrick and Pyle in experiments on ink-blots found a tendency of the imaginativeness to decline with increas- ing age. Such a tendency in this test would be explicable on the assumption that children grow more critical of their perceptions as they grow older. On the other hand, children acquire a richer fund of experience as they grow older and therefore ought to be able to "see" more in the blots. The "nascent stages" for imagination differ, of course, with the type of imagina- tion concerned. In an earlier experiment in suggesting perspectives, the most suggestible age between 9 and 16 seemed to come at 12 and the least at 16. 1 5. The extent of the individual differences in this test is indicated by the following figures, showing the lowest and the highest number of "imagina- tions" given for each ink-blot (sittings 2 to 5): 2 3 > 4 5 For the epileptics 0 to 9 0 to 6 0 to 10 0 to 8 For the normals. 0 to 14 0 to 12 0 to 15 0 to 14 The individual differences in each age for the normals, based on the averages for sittings 2 to 5, are as follows: Age.. 7 8 9 10 11 12 13 14 15 16 17 Difference 6.5 5. 3.7 7. 5. 12.2 5.5 8.4 5.8 3.7 2.8 6. This does not seem to be a very valuable test for measuring the (imaginative) capacity of children of various ages nor of measuring extent of improvement with time: first, because the performance in the test does not correlate significantly with age; second, because the different ink-blots are of unequal degree of suggestibility; and, third, because the trait concerned may be more specific, in spite of what we have said on page 125, than most of the other traits which we have here measured. We need to determine the degree of correlation between "imagination" and general intelligence. 'J. E. Wallace Wallin. Optical Illusions of Reversible Perspective, 1905, 213—232. X. WORD CONSTRUCTION Building Words from Six Supplied Letters: II to 15. 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS Six letters in 10 point type were printed at the top of the test blanks, which measured 41/4 x 7x/2 inches (see below; 16 was not used in this experi- ment). In attempting to determine the best combination of letters to use for the different blanks we made preliminary tests with twelve different sets of six letters (three vowels and three consonants in each set) on three col- leagues (physicians). The experimenter also tested himself. These prelim- inary tests indicated that there was considerable difference in the difficulty of the different sets, from the standpoint of the number of words which were con- structed in the same amount of time. Four sets seemed to be almost in a class by themselves, namely, aeobmt;eairlp (both of these from Whipple), o e a s d n (from Terman). and i a e h t f. for there were more words written with these than with any other com- bination of letters. And yet we cannot conclude that these four are of equal difficulty, for more words were written with the first two sets of letters (some doing better in the first and some in the second), than with the last two sets. It will be observed from the key supplied in the third section that more words can be constructed from the combination e a i r 1 p than for a e o b m t. and more words were actually made by our normal sub- jects with the former set of letters. Curiously, no one regarded e a i r 1 p as the easiest combination. Two said that the easiest combination was a e o b m t. Two did not know which was the easiest. One said the combination i a e h t f was the hardest, and two the combination o e a s d n. These judgments, however, were not always borne out by the number of words which the different subjects constructed. We did not deem it advisable to include any of the other sets which we investigated, and have therefore been obliged to use two combinations twice, as shown below: a e o b m t II a e o b m t 12 e a i r 1 p 13 e a i r 1 p 14 - - 132 Normal and Epileptic School Children 133 0 e a s d n 15 1 a e h t f 16 We must have available a sufficient number of sets of standardized letter combinations before we can hope reliably to measure progress in word-build- ing capacity. 2. NATURE AND PURPOSE OF THE TEST Success in this test depends upon a number of factors: extent of vocab- ulary, ability to spell (and write) correctly, fertility of invention or resource- fulness in combining the letters in a great number of ways, and ability to keep the attention concentrated on the problem. Other things equal, the person who has the largest vocabulary and who can spell correctly will score highest in the test, because he will recognize more genuine words, and will thus be able to reject imaginary and misspelled words. Some of our subjects wrote a considerable number of imaginary words. Of course, for the time spent in writing these words they received no credit. On the other hand, when no demerits are imposed for the wrong words written a person of lim- ited vocabulary who is very fertile in combining the letters in a great variety of ways might overcome the advantage possessed by the one who has a rich vocabulary, for some of the numerous words written will of necessity be cor- rect. Some words will be correct as a mere matter of chance. It is evident that the person who possesses a good vocabulary and, at the same time, is very resourceful in associating the letters in varied ways, and who is able to keep his attention highly concentrated will excel in the test. Besides testing resourcefulness, ingenuity, and ability to hold attention on a task, we also test the individual's "combining" activity in this test — that is, the ability of the individual to combine isolated parts into a totality. According to Ebbinghaus, the combining of elements into a whole is of the very essence of intelligence. From these several points of view we seem justified in classing the word- construction test as a test of intelligence. 3. DIRECTIONS FOR GIVING THE TEST Time: Exactly three minutes. Place the blanks with the letters upside down on the desks, with instructions to the pupils not to turn them until they are requested to do so. Instructions to Pupils: "On the other side of the blanks you will find at the top six letters. I want you to make as many words as you can out of these six letters. Put them together into as many different words as possible. You must not use more than six letters in any word; you must not use any 134 The Measurement of Mental Traits in other letters than those given ; and you must not use the same letter twice in the same word. Let me show you what I mean with these letters: u i a d c n (Place on board). From these letters we can spell: cud, cad, can, Dan, Dina, ai, aid. But not: nina, or nun, or naiad." (It is well to emphasize that the words coined must be recognized words and not mere inventions.) "Now, ready!" "Turn!" "Now, stop! Turn! Sign your names." Scoring: Record the number of words correctly constructed. The number of words wrongly constructed may be noted, and a tabulation of the words may also be made. The imposition of a penalty for wrong words would probably discourage guess work. The following key indicates the words which may be constructed from three of the combinations. We have taken pains to make the lists unusually complete for the first two sets of letters. The list for the latter set is incom- plete. We have included obsolete and dialectic words, abbreviations and words spelled according to the reformed spelling, for the first two sets of letters. These unusual words are given in italics. Lists without the unusual words have been provided by Whipple for a e o b m t and e a i rip. A E O B M T a, ab, Abe, abet, abt, am, ambo, amt, at, ate, atmo, atom, ba, bam, bat, bate, be, beam, beat, bema, bet, beta, bo, boa, boat, bom, boma, bot, bote, ea, earn, eat, eb, em, eta, ma, mao, Mae, Mab, mat, mate, me, meat, met, meta, mo, moa, Moab, moat, mob, moe, mot, mote, o, oat, ob, obe, obt, om, ta, tab, tabe, tarn, tambo, tame, tea, team, tema, to, toe, Tom, tome. E A I R L p a, ai, aiel, ail, aile, air, aire, al, ale, alp, ape, apl, April, ar, are, Ariel, aril, ea, ear, earl, el, ela, Eli, epi, er, era, eria, I, il, He, Ira, ire, irpe, la, lair, lap, lare, le, lea, leap, Lear, lep, lepra, lerp, It, liar, lie, lier, lip, lira, lire, lirp, pa, paie, pail, pair, pal, pale, paler, pali, par, pareil, parel, pari, pea, peal, pear, pearl, pel, pela, per, peril, pi, pia, pie, piel, pier, pil, pile, pilar, piler, pire, plea, plie, plier, prial, prie, Ra, Rae, rail, rale, rap, rape, re, Rea, real, reap, rei, rep, rial, ril, rile, rip, ripe. o e a s D N a, ad, aden, ae, an, and, as, e, ean, ed, end, Eos, da, Dan, dane, de, dea, dean, dzas, den, do, doe, does, don, done, dose, na, nae, ne, nea, Ned, no, nod, nose, o, od, ode, on, one, one's, os, sad, sade, sae, san, sand, sane, sea, sead, sed, sedan, sen, send, so, sod, soda, soe, son. Normal and Epileptic School Children 135 4. TABLES TABLE 31 Building Words from Six Supplied Letters: 11—5 Sittings 1 and 2: a e o b m t Sittings 3 and 4: e a i r 1 p Sitting 5: o e a s d n Normal Children No. Ae. Sx. 1 No. 2 No. 3 No. 4 No. ft 5 No. 1-2 Av. No. 3-4 Av. No. J Ave 1-5 2 7 B 2 5 2 0 80 1 5 2 0 1 33 2 0 2 2 1.7 .77 2 0 2 Q 3 0 2 0 2 00 4 7 0 1 75 3 5 3 2 5 5 2.50 4 1 4 7 Av. 2 7 2 0 1 40 2 7 4 5 1 61 2 7 2.7 3 6 1.92 3 0 7 3 B 4 1 4 1 1 19 5 1 5 1 99 4 5 4.1 4 8 1.21 4 5 3 g 4 0 6 3 1 41 2 ti 2 0 1 05 3.6 4.8 2 3 .52 3 6 10 8 Av. 4 4 4 1 25 4 4 4 2 1 01 4 2 4 3 4 5 1 02 4 2 3 9 B 1 3 5 ? 50 3 6 3 3 1 22 2 6 2. 3.5 1.75 2 7 4 9 G 4 7 4 0 1 40 7 0 4 7 75 5 7 4 3 6 1 1 70 5 1 7 9 Av. 3 1 3 B 2 24 5 3 4 1 .99 4.4 3 3 4.8 1.73 3.6 4 10 B 8 5 9 6 1 32 11 3 11 5 1 20 9.7 9.4 10 6 1.22 10 4 3 10 G 6 6 6 0 84 5 33 7 3 1 43 7 7 6 3 7 6 1 42 6 6 7 10 Av 7 7 7 s 1 08 8 3 8 3 1 32 8 8 8 1 9 3 1 31 8 7 3 11 B 9 6 9 0 86 9 0 13 0 1.75 9.6 9.3 11.0 1 28 10.6 8 G 8 2 11 3 1 57 8 8 9 8 1 13 8 0 9 g 9 3 96 9 6 11 11 Av. 8 6 10 6 1 38 8 9 10 7 L32 85 9^6 9^6 1.06 9.8 5 12 B 8 8 10 0 1 23 9 4 12 7 1 21 8.8 9 4 10.3 1.10 9 5 3 12 G 7 6 12 3 1 72 11 6 12 0 1.19 9.3 10 0 11.6 1.07 10 6 8 12 Av. 8 2 10 8 1 43 10 2 12 4 1.20 9.0 9 6 10 6 1.09 9 9 4 13 B 9 S 16 6 1 68 13 2 15 5 117 14 7 11.6 14 3 1.43 13 4 2 13 G 10 0 9 0 92 10 5 12 0 1.30 12.0 9.5 11 2 1.12 10.7 6 13 Av. 9 7 13 6 1 39 12 3 14 .3 1 21 13 8 10 9 13.3 1.33 12.6 4 14 B 9 0 12 6 1 27 11 0 12 6 1.28 8.0 10. 0 11 7 1.16 10.3 7 14 G 8 4 15 1 1 91 13 2 14 3 1.16 13 1 11.8 13. 5 1.15 12.4 11 14 Av. 8 6 14 4 1 72 12 2 13 1.19 11.6 11 1 12 8 1.15 11.6 2 15 B 16 0 17 0 1 04 19 0 19 0 1.10 12.5 16 5 18.5 1.13 16.0 2 15 G 16 0 14 0 86 18 0 16 0 89 7.0 15.0 14. 5 .96 14.2 4 15 Av. 16 0 15 5 98 18 3 17 5 .96 9 7 15 7 16.3 1.05 15 1 1 16 B 15 0 16 0 1 06 14 0 13.0 15 5 14.0 90 14.5 3 16 G 7 6 12 5 1 39 10 3 10 0 "95 14 0 9.0 10.1 1.33 10.8 4 16 Av. 9 5 13 6 1 28 10 3 11 0 95 13.7 10.6 11.1 1.22 11.7 1 17 B 13 0 18 0 1 38 18 0 21 0 1.16 15.0 15.5 19 5 1.25 17 0 2 17 G 16 0 16 0 1 06 14 5 17 0 1.24 11.0 16.2 15.7 1.07 14.9 3 17 Av. 15 0 17 0 1 22 15 6 18 3 1.23 11.3 16.0 17.0 1.13 15.6 Ave S 7 B 9 2 1 35 8 7 10 5 1 21 8.2 8.5 9.6 1 24 8 8 Ave S 8 0 10 6 1 47 9 5 10 1 1 15 9 0 9.2 9.8 1.19 9 4 Ave S &S 7 9 9 9 1 41 9 1 10 3 1.18 8.6 8.9 9 7 1.21 9 1 136 The Measurement of Mental Traits in TABLE 32 Building Words from Six Supplied Letters: 11—5 Sittings 1 and 2: a e o b m t Sittings 3 and 4: e a i r 1 p Sitting 5: o e a s d n Epileptics Su. Ae . c„ bx . 1 No. 2 No. 3 No. 4 No. - 5 No. 1-2 Av. No. 3-4 Av. No. Ave 1-5 81 10 Boy 3 3 2 3 3 1 nn L . O 102 10 Girl 11 6. g' 1 00 j J g □ D fi ^ 104 10 Girl •j g 85 2 4 2 00 E D . 0 . j 0 . .40 i4 fi 4.0 Av 9 g' . OO 4 5' 1 50 Q O . a 0 . . OU 0.4 ' 'AA' ' yy * , 11 3. 6 2. GO 6. 6. 4.5 6. 1 33 5.2 79 12 Boy 1 _ 1 00 0 ] j 1 1 . c .0 . ou A . 0 84 12 Boy g' g' 1 00 g g' 1 00 c u . fi 1 nn 0.0 91 12 Boy 0- 0 0 ] 0 Q 5 2 Av 2 .3 2 3 1 .uu 0 £ 2 6 1 00 1 fi £ -0 7S 9 9 6.4, 108 12 Girl 0. 4 n 0 2 0 n a 93 13 Boy 2 ' 1 _ 50 5 0 0 £ ■ 1 c Q . u i i . 86 13 Boy 0 '0 2 1 ' .50 Q 5 15 0 - g Av 15 . ZD i 5 RA . 3U 1 1 1 . 7 1 ^ 1 . 0 Q 97 13 Girl 4 4 ' 1 00 1 g 4 3 7 77 14 Boy 5' 4 80 g g 1 33 g 4 5 7* 1 ss 1 . DO ^ fi O 0 89 14 Boy 4 q" n 4 4 1 00 9 4 9 no 1 R Av. 4^5 2. .40 5' 6^ 116 6 5 3.2 5i5 1.77 4.8 "80 ' is' Boy 4- 5. 1.25 2. 2. 1.00 5. 4.5 2. 44 3.6 82 15 Boy 9 9 00 7 9 1 28 B 3 . ■ 1 AD 0- u 9 1 • 1 . 1 . .5 .50 .8 82 15 B 0. 0. 2. 1. 2. 1. .5 .50 1.0 83 15 B 0. 0 0 0. 0. 0. 0. 0. Ave. 0. 0. 1.3 G 1. .6 .3 ".50 .6 103 is' G 0. 0. 0 0. 0 0. 0 0. 78 16 B A U. n U - 1 n u . ft E .0 U. .uu .2 88 16 B 2. 3. 3. 3. 3 2.5 3 1 20 2.8 85 16 B 3. 1. 3. 1. 2. 3. 1. .33 2 Ave. 1.6 1.3 2.3 1.3 1.6 2.0 1.3 51 1.6 96' 16 G 0. 0. 0. 0. 0. 0. 0. 0 98 16 G 2. 0. 0. 0. 1. 0. ".00 0.5 106 16 G 0. 0. 0. 2. 2. 0. 1. .8 100 16 G 0. 0. 0. 0. 0. 0. Ave. .5 0. 0. .5 .5 2 ".3 .00 .2 87 W B 0. 0. 0. 0. 90 17 B 0. 6. 0. 0. 0.' 0. 0. 0. Ave 0. 0. 0. 0. 0. 0. 0. 0. 95 G 0. 0. 0. 0. 0. 0. 105 18 G 2. 0. o! 0. 0. 1. 0. .00 .4 107 21 G 0. 0. 0. 0. 0. 0. 0. 0. 101 24 G 0. 0 0. 0. 0. 0. 0. 0. Ave. Boys .9 .3 1.2 .9 1.2 1.0 .6 .51 .9 .7 0. .3 .4 4 .5 .2 .20 .3 Ave. Boys & Girls — 8 .1 .8 .7 .8 5 .4 .40 6 5. RESULTS (See Tables 33 and 34) The average number of sentences constructed in all the sittings (last column in the tables) was 3.8 for the normals and .68 of a sentence for the epileptics. The epileptics built only 17.9% as many sentences as the normals. The epileptics did only 45% as well as the youngest group of the normals and only about 10% as well as the oldest group of normals. Eight of the epileptics were unable to construct a single sentence in five sittings, six in four sittings, four in three sittings, six in two sittings, and two in one sitting. In other words, 86% of all the epileptics failed in one or more sit- 144 The Measurement of Mental Traits in tings. Seven of the normal pupils failed completely in all five sittings, one in three sittings, two in two and sixteen in one. Thirty-four per cent of the normal pupils failed in one or more sittings. Of course, this large difference may not be due wholly to the poverty of the epileptics' "linguistic imagina- tion," or their inability to construct the sentences, but also to the difficulty of writing the sentences. Both the epileptic and the normal pupils wrote more sentences with the nouns than with the verbs. The difference between the averages for the sentences in sittings 1 and 3 (nouns) and 2 and 4 (verbs) amounts to .9 of a sentence for the normals and .1 of a sentence for the epileptics. The differ- ence is slightly greater, both absolutely and relatively, for the normal than for the epileptic children, although this does not seem to be the case from the index of loss in f '\Z$. According to the indices, the normals did 20% poorer with the verbs than with the nouns, while the epileptics did 60' i poorer. But many of the indices could not be computed for the epileptics, because of the negative scores, whence the epileptics' indices are not very reliable. The normal boys constructed only 74% as many sentences as the normal girls, or 1.1 sentences less each, based on the averages for all five sittings (last column), while the epileptic girls wrote only 40% as many sentences as the epileptic boys, or .54 of a sentence less each. The normal girls did better than the normal boys in every sitting, the greatest difference between the gen- eral averages amounting to 1.4 sentences in sittings 1 and 5 (nouns) and the least difference to .8 of a sentence, in sitting 1. On the other hand, the epileptic boys surpassed the epileptic girls in every sitting, the greatest differ- ence being .9 of a sentence, in sitting 3. and the smallest difference .2 of a sentence in sitting 1. Based on the sex averages for the different ages for all the sittings (last column), the girls excelled in seven ages and the boys in four among the normal pupils and the boys in three ages and the girls in two among the epileptics, the averages being the same in one age. Based on the age averages for all the sittings (last column) for the normal children, there was an increase in the number of sentences written in six of the ages and a decrease in four, namely in ages 8, 12. 16 and 17. in which, respectively, 72%, 62%, 75% and 100% of the pupils were classed as dull and average. The difference between the averages of the youngest and the oldest group amounts to 4.9 sentences. The highest score, however, was made by the fifteen-year group and the lowest by the eight-year-olds. 6. CONCLUSIONS 1. The normal pupils were more than five times as efficient as the epi- leptics in this test of sentence construction. We imagine that the epileptics would have done somewhat better relatively had the responses been verbal Normal and Epileptic School Children 145 instead of written. The test was too difficult for some of our normal and epileptic cases. Binet placed a similar test in age X. Our normal ten- year-olds do more than twice as well as the children in any of the younger ages. 2. Ihe individual variations were considerable in the ability to con- struct sentences, under the conditions of this test. Among the epileptics the number of sentences varied as follows in the different sittings: in 1, from 0 to 4; in 2, from 0 to 3; in 3. from 0 to 3; in 4, from 0 to 3; and in 5, from 0 to 4. Among the normals the variations for children from 7 to 17 were as follows: in 1. from 0 to 10; in 2, from 0 to 9; in 3, from 0 to 13 ;in 4, from 0 to 10; and in 5. from 0 to 12. The difference between the poorest and the best records in each age among the normals, based on the average of all five sittings (1-5), was as follows: Age.. 7 8 9 10 11 12 13 14 15 16 17 Difference. 2.8 3.2 4.0 1.6 4.6 2.9 4.4 3.4 3.6 3.8 2.4 3. Both the normals and the epileptics wrote more sentences with the nouns than with the verbs, a result in harmony with Sharp's findings, al- though she says that the sentences with the verbs were better in quality. 4. The normal girls did appreciably better than the normal boys, while the epileptic boys did decidedly better than the epileptic girls, in the con- struction of sentences. 5. The ability to construct sentences increases with ascending age among the normal children. The oldest group of pupils built more than four times as many sentences as the youngest group. The exceptions which we find are to be attributed to a preponderance of dull and average pupils in certain ages and to the fewness of the subjects. 6. Our experience with this test, both when given to groups of children and when given clinically (as in the Binet scale), has uniformly pointed to its value as a test of intelligence. In clinical examinations we have found verbal responses preferable to written responses, especially when the subject is asked to make only one sentence, as in the Binet test. The test, however, is too difficult for seven or eight-year-old children. M T 10 XII. RECOGNITION Recognition of Illustrated Post Cards: K4 1. REPRODUCTION AND DESCRIPTION OF TEST MATERIALS The test materials consisted of two sets of illustrated post cards, pasted on white cardboards, 22x28 inches; first, the stimulus set, No. 1-A, which contains 10 post cards; and, second, the recognition set, No. 2-A, which con- tains twenty post cards. Ten of the post cards on 2-A are exact duplicates of the cards on 1-A. The imperfections which appear in the photographic reproductions, from which the cuts on pages 147 and 148 are made, did not appear on the post cards themselves. On the other hand, ten of the post cards on 2-A are new. The post cards on 2-A were numbered consecutively from 1 to 20, while the cards on 1-A were not numbered. The post cards were made from unusually clear prints and, together with the numbers, were easily visible from any seat occupied by the examinees. Parallel sets of cards, 1-B and 2-B, were constructed, but we were unable to use these. Cardboard 1-A (the stimulus card) was exhibited for 20 seconds at the be- ginning of the sitting (before test A-4 was given), while 2-A (the recognition card) was exposed for 90 seconds at the end of the sitting (after test L), about an hour later, in accordance with the procedure described in section 3, below . 146 The Measurement of Mental Traits Normal and Epileptic School Children 149 2. NATURE AXD PURPOSE OF THE TEST In this test the subject is asked to examine very carefully the cards on 1-A, so that he will have them thoroughly "fixed in his mind." After the proper interval he is asked to scrutinize the cards on 2-A, and determine which of the cards appeared on 1-A. Obviously success in the test i3 de- pendent upon the subject's ability (1) to observe accurately the character- istics of each post card, to note points of difference and resemblance between them, particularly the cluster of attributes which makes one card differ from any other card; (2) upon his ability to retain the impressions received from the inspection of the cards with their associated images; and (3) upon his ability to identify the original cards when later presented in a different set- ting among an equal number of strange cards, some of which resemble the original cards more or less closely. Incidentally, success may also depend on the ability to perceive and record correctly the numbers of the cards. Such a process of conscious identification of the individual's present experience with his past experience is known as "recognition," and is said to rest upon a feeling of "familiarity." It is evident that the more the cards on 1-A are like the new cards on 2-A, or the less distinctive the cards are on 1-A, the more difficult will be the process of recognition. Thus the second card (flower) on 1-A, can easily be recognized on 2-A, because the flower stands out distinctly and no other card on 2-A closely resembles this card. On the other hand, the houses on post cards 9 and 20 might very easily be confused. The difficulty of the test can be very greatly increased by using cards which are quite similar in appearance, and, also by lengthening the interval of time between the original impression and the recall. In the fourth sitting the interval was only one hour (we shall discuss these results under the designa- tion of "immediate recognition"), while in the fifth sitting it was one month ("deferred recognition" of the cards displayed one month earlier). Success in the test, then, depends upon the individual's degree of active attention, keenness of observation and power of retention. The strength of these factors, in the large, probably varies with general intelligence. 3. DIRECTIONS FOR GIVING THE TEST Time: 20 seconds for the exposure of the stimulus set (1-A); 90 sec- onds for the exposure of the recognition set (2-A), during which time the subjects make their records on plain sheets of paper. Directions: At the beginning of the experimental period the experi- menter, without further notice, made the following announcement: "I am going to show you some post cards on the other side of this card- board (stimulus set, 1-A). I want you to look very carefully at the cards, so that you will get a good impression of each one of them. Try to -fix them 150 The Measurement of Mental Traits in in your minds as well as you can. I will only allow you a short time in which to look." "Now, ready!" (Turn the cardboard rapidly and expose the cards for 20 seconds.) The following statement was then made: "A little later I shall ask you to do something with the cards." At the end of the hour, the experimenter proceeded. "Now I shall show you a set of cards which includes all the cards which I showed you before, and also a number of new cards. Each post card is numbered. The number is written above the card. What I now want you to do is to look the cards over carefully when they are exposed and write down the numbers of all the cards which I showed you before. If card 1 was shown before, write the figure 1; if card 2 was shown, write the figure 2. etc. Do not write the figures of cards which were not shown before." "Now, ready!" "Now, stop. Turn your papers. Sign your names." Scoring: For percentage of efficiency, score 10% for each post card correctly recognized. Record the number of cards mistakenly recognized. A proper deduction may be made for mistaken recollections, but this was not done in this experiment (except when all the numbers were written). The announcement that a deduction will be made for all wrong numbers written would probably tend to make the subjects more cautious and lessen the tendency to write all the figures or a large number of figures at random. It is evident that when no deduction is made for wrong numbers a very poor child could make a perfect score by writing all the numbers. We had to give a zero score to ten of the epileptics and four of the normals (three in the second and one in the fourth grade), because they wrote all the numbers. A considerable number of pupils wrote several wrong numbers, although we emphasized the fact that only the numbers of the cards recognized as having been shown before should be written. Most of the pupils who wrote wrong numbers, however, probably made honest mistakes. Another method of discouraging the tendency of writing too many num- bers at hap-hazard is to tell the subjects that not more than ten numbers may be written. This was not done in this experiment. One of our epileptics wrote 11 numbers, one 12, and one 13, all in the deferred test. Three of the normal pupils wrote 11 numbers, three 12 and one 13. in the immediate test, and four wrote 11 and one 14 numbers in the deferred test. It is evident that these pupils reaped an unjust advantage. The number of cards wrongly remembered is recorded in the tables in the "W" column. Normal and Epileptic School Children 4. TABLES TABLE 35. Recognition of Illustrated Post Cards: K4 151 Normal Children Epileptics No. Age Sex 4 No W. 5 No. W '. u . 1 Age Sex 4 No. W 5 fo No W '. 2 7 B 30.0 2 25 0 4 81 10 B o. 80. 4 2 7 Q 70 0 1 ' 50.0 4 102 10 G 10. 1 o' 0 4 1 Ave. 50 o 1.3 37 5 4 104 10 G 40 ' 3 20. 0 7 8 54 3 2 47 1 3A Ave 25! 2 10 3 8 56 2 3 6 56 0 3 6 79 12 B O.l 0.1 10 8 Ave 55 .0 2.5 50 0 35 84 12 B 80 ^ 1 80 3 3 9 43 3 13 46.6 3 6 i 91 12 B 0.1 50 '. 4 4 9 G 42 5 2 3 42 5 2.6 ! . . 26 '6 43.3 3.5 7 9 Ave. 42 8 2 2 44.2 3 2 i 108 12 0 1 0 1 4 10 77 5 1 60 0 3 i 93 13 B 0 1 0 1 3 10 G 70.0 16 g6 6 2 3 86 13 B 0 1 10. 1 7 10 Ave. 74 S 13 62 '8 2 7 5^ 3 11 50 .0 2 30 .0 3 3 77 ' ik' 90. 0 70 ' 2 8 11 G 62 5 1 7 48 5 1 8 89 14 B 40 2 40 6 11 11 Ave. 59 1 18 43^0 2 3 Ave. 65! 1 55^ 4 5 12 B 67 5 1.5 62 5 3 7 80 15 B 30. 2 60. 4 3 12 G 76 6 i[ 6C 2 6 82 B 70 3 8i) 5 a 12 Ave. 71 4 13 61 4 3 3 15 B 30 2 10 0 4 13 57^5 2.7 55^0 3.3 Ave 43^3 2.3 50. 3 2 13 G 60.0 10 60 0 2 0 103 is G 0 0 0 0 6 13 Ave 58 3 2.1 56.6 2 8 78 16 B 10. 0 0. 4 14 B 76 6 .6 70. 1.5 88 16 B 40. 1 "i0 5 7 14 G 74.3 .8 61.4 1.5 85 16 B 0.1 11 14 Ave. 75.0 .8 63 3 1.5 Ave. 16.6 ".5 16.6 2 15 B 80.0 15 70.0 2.5 96' 16 G 0. 0 0. 0 2 IS G 80 0 1 65.0 3.5 98 16 G 10. 3 0. 4 15 Ave. 80.0 1.2 67 5 3.0 106 16 G 0.1 30. i 1 16 B 80.0 0. 60.0 1 100 16 G 40. 5 0. 2 3 16 G 86.0 1.6 63 3 3. Ave. 12.5 2.6 7.5 .7 4 16 Ave. 85.0 1.3 62.5 2.5 87 17 B 0.1 0.1 1 17 B 70.0 3 40 0 4. 90 17 B 0. 0" 0. 0 2 17 G 85.0 0. 75. 2.5 95 17 G 0. 0 0. 0 3 17 Ave. 80. 0 I. 63.3 3. 105 107 18 21 G G 0.1 0.1 0.1 30. '2 Ave. Boys 60.8 1.7 51 5 3.1 101 24 G 0. '6" 0. 0 Ave. Girls . 67.9 1.5 57.3 2.4 Ave. Bovs & Girls 64.6 1.6 54.6 2.8 Ave. 3oys Girls. 24.3 8.3 1.2 33.1 2.8 Ave. 1.5 6.6 .5 Ave. Boys & Girls; 17.5 1.3 21.7 1.7 ino. w indicates tne number ot post cards wrongly recognized ■Wrote all the numbers consecutively, hence given a zero score. None of these subjects was considered in the column containing the wrong numbers. 5. RESULTS (See Table 35) IMMEDIATE RECOGNITION The normal pupils correctly recognized on the average six and a half cards (which gives a score of 64.6%), while the epileptics' average was one and three-fourths cards (which gives a score of 17.5%). In power of visual recognition the epileptics did only 27% as well as the normal children. They did only 35% as well as the youngest group of the normals and not quite 22% as well as the oldest group of normals. Fifteen epileptics failed entirely (ten of these, however, because they wrote all the numbers), and four of the normal pupils (all of whom wrote all the numbers). 152 The Measurement of Mental Traits in It is probable that the epileptics' average is slightly too low, because of the losses incurred through the ten pupils who recklessly wrote all of the numbers and who were given zero scores. It is evident that some of these subjects must have been able to recognize some of the cards. This is indi- cated by the fact that the average for the epileptics in the deferred recogni- tion test, which was given one month later, is actually higher. The pupils were then explicitly told that many had written all the numbers in the previous test and that this must not be repeated. Two normal pupils (girls, aged 14 and 16) had all ten cards correct. No epileptic made a perfect score, but one recognized nine cards correctly and one eight cards. The normal boys recognized 89% as many cards as the normal girls, or seven-tenths of a card less each. On the other hand, the epileptic girls recognized only 34% as many cards as the epileptic boys, or about one and a half cards less each. Among the normal pupils the girls excelled the boys in seven ages, the boys excelled in three ages, while the scores were the same in one age. Among the epileptics the girls did better in only one age, the boys in three ages, while the scores were equal in the remaining age that could be compared. The ability to recognize the cards increased in six ages, while there was a decrease in four ages, namely, in ages 9, 11, 13 and 17, in which, respect- ively, 57%, 72', , 50% and 100% of the pupils were rated as average and dull. The average number of wrong numbers recorded was 1.6 for the normal pupils and 1.3 for the epileptics, a difference of .3 of an error. The normal pupils made 25% more mistaken recognitions than the epileptics, while the epileptics made only 27% as many correct recognitions as the normal pupils. It must not be forgotten that we are not including the records of the errors of those subjects who wrote all the 20 numbers. Had these been included the epileptics' average number of errors would have been much larger. Moreover, the number of wrong cards given should be considered in relation to the number of right cards given. The average number of wrong recogni- tions made by the normals amounted to only 24% of the average number of right recognitions (6.46), while the average number of errors made by the epileptics amounted to 74% of the average number of cards they had right (1.75). Relatively to their degree of efficiency the normals made fewer errors than the epileptics. The normal girls made 88% as many false identifications as the normal boys, or .2 of one error each less than the boys, while the epileptic boys made 80% as many as the epileptic girls, or .3 of one error each less than the girls. Among the normals the girls made less errors than the boys in six ages, while Normal and Epileptic School Children 153 the boy? made less errors than the girls in five ages. Among the epileptics comparison was possible only in three ages. The average number of errors increased in five ascending ages and de- creased in five ascending ages among the normal pupils. DEFERRED RECOGNITION Without prior warning the recognition cardboard, 2-A, was shown at the beginning of the fifth sitting, and the pupils were asked to write the numbers of all the cards they were able to recall as having been shown one month earlier on the stimulus cardboard 1-A (which, of course, was not again ex- posed). They were allowed the same amount of time for writing as before. As has already been indicated, they were specially cautioned not to write all the numbers, but only the numbers of the cards which they beleived they saw on the other cardboard (1-A). The normal pupils on the average correctly recognized about five and a half cards (scoring 54.6%), or one card less than in the immediate recogni- tion. The loss in the ability correctly to recognize the cards during the lapse of one month thus amounted to 16%. The epileptics recognized on the average a little over fwo cards (scoring 21.7%), or almost half a card more than in the immediate recall. The epileptics gained 24% , while the normal pupils lost 16%. In the immediate recall the epileptics did only 27%> as well as the normals, but in the deferred recall they did 39% as well. The gain made by the epileptics, however, is, we believe, spurious. They did better in deferred recognition, not because of any increase in the power of recogni- tion, but partly because they were specially cautioned not to write all the numbers, which would deprive them of all credit, but only the numbers of the cards actually exhibited on cardboard 1-A. Subjects 91, 86, 106 and 107 profited by this advice, while 79, 108, 93, 85, 87 and 105 did not, includ- ing 78, who wrote all the numbers, although he did not do so the time before. One subject, 81, made no score in the immediate recognition test (returning a blank page), while he got eight cards correct in the deferred recognition test. It will be seen, however, that he wrote twelve numbers. Two other subjects, 82 and 84, wrote over ten numbers. Two of these three subjects made a higher score in the deferred test. The greatest number of cards correctly recognized was eight, namely, by nine normal pupils and three epileptics — all of the epileptics, however, wrote too many numbers. The normal boys did almost 90% as well as the normal girls, recognizing on the average fifty-eight hundredths of a card less than the girls. The girls excelled in seven ages and the boys in four. We found the girls superior in the immediate recall approximately in the same ratio. On the other hand, the epileptic girls did only about 20% as well as the epileptic boys, recogniz- 154 The Measurement of Mental Traits in ing on the average two and a half cards less than the boys. The epileptic boys excelled in four ages, while the scores were the same in the other age which could be compared. Relatively the epileptic girls did markedly better in the immediate recognition than in the deferred recognition. The ability to recognize the cards increased in six ages and decreased in four ages among the normals, just as in the case of the immediate recall. Three of the ages showing a decrease were the same as in the immediate recognition, while age 16 showed a decrease in the deferred recall instead of age 17. The normal pupils wrote on the average 2.8 wrong numbers, or 1.2 more than in the immediate recall. In other words, they made 75% more errors in the deferred than in the immediate recall. The epileptics averaged 1.7 wrong numbers, or .4 of an error more than in the immediate recall. They made 30% more errors than in the immediate recall. The errors of the normals amounted to 51% of the score they made in the deferred recall (5.46), while the errors of the epileptics amounted to 78% of their score (2.17). Relatively to the degree of efficiency which they showed, the normal pupils made less errors than the epileptics, but the epileptics did better relatively in the deferred recall than in the immediate recall, probably due to the greater care taken and to the inclusion of records which could not be con- sidered in the earlier test. The normal girls made 77% as many mistakes as the normal boys, or on the average .7 of one error less. The normal girls made less errors in six ages than the normal boys, the boys made less errors than the girls in three ages, while the results were the same in two ages. Relatively to the boys, the normal girls made less errors in the deferred than in the immediate recall. The epileptic girls made only 18% as many errors as the epileptic boys, although in the immediate recall they made 20% more errors than the boys. Since, however, only three girls made a score in the deferred test, while eight either wrote nothing at all or wrote all the numbers (which thus were re- jected) in the immediate recognition test, the comparative results are not reliable. INTROSPECTIONS ON THE DEFERRED RECOGNITION The two sections containing the oldest normal children were asked to indicate how they were able to recognize the post cards which had been pre- sented a month earlier on 1-A, whether by recalling the cards themselves— the pictures on the cards — or by recalling the numbers they wrote a month earlier, or by visual imagery, or by some other method. Thirty-nine answered that they remembered the cards themselves or the picture on the cards (one remembered only parts of the cards), while eight recalled the pictures as well as the numbers, or some of the numbers they had written. Normal and Epileptic School Children 155 The pupils in the oldest section of normal children were 'also asked to state whether or not they were certain respecting the correctness of their recognitions. Five stated that they were uncertain. But two of these who said their recognitions were "guesses" had six right and only one wrong and eight right and only one wrong, respectively. One had four right and three wrong, one five right and two wrong, and one six right and one wrong. The records of four of these pupils were much better than could have been in- ferred from their introspections. Six said they were uncertain of some of the post cards. Of these six three had eight right and two wrong, one had six right and four wrong, one six right and three wrong and one four right and two wrong. One who was certain of most of the cards had seven right and three wrong. One who was certain of four or five had five right and three wrong. One who was certain of only one card had four right and four wrong. One who was certain of those he gave in his answers had six right and two wrong. Of four who were certain of their judgments, two had eight right and two wrong, one had seven right and three wrong and one had five right and none wrong. On the whole, there was a considerable degree of correlation between the subjective certainty or uncertainty and the empirical results. 6. CONCLUSIONS 1. According to the empirical result the epileptics did only 27 '% as well as the normals in this test. This figure, however, does not correctly represent the epileptics' relative inferiority in the immediate power of recognition, owing to the fact that many subjects were given too low a score (zero) because they wrote all the numbers. At least some of these subjects would have made some score had they properly followed the instructions. 2. The ability correctly to recognize the cards after the interval of a month was appreciably weakened in the normal children — a loss of 17% in efficiency — while it was increased in the epileptics (by 24%)- We have given reasons, however, why the epileptics' gain after a month's interval cannot be considered to represent a genuine improvement in the power of recognition. 3. The normal girls excelled the normal boys, while the epileptic boys decidedly excelled the epileptic girls, in the strength of recognition, both im- mediate and deferred. 4. The power of recognition, both immediate and deferred, tends to increase slightly with ascending age; the exceptions, which are relatively numerous, are probably explained by the limited number of subjects and the unequal distribution of pupils of different degrees of ability in the different ages. The difference between the oldest and the youngest group of children amounted to 30% in the immediate recognition and 25.8% in the deferred recognition, a difference of two and a half cards. The older pupils did 1.6 156 The Measurement of Mental Traits in times as well 'as the youngest pupils in both the immediate and in the de- ferred test. 5. The absolute number of mistaken recognitions was larger for the normal than for the epileptic children, both in the immediate and in the deferred recognition. The reason that the epileptics made less errors is due to the rejection of the records of all the epileptics who wrote all the numbers, and to the fact that some epileptics did not write any numbers, while some wrote only four or five numbers altogether. However, when the number of wrong recognitions made is compared with the number of correct recognitions, the normal pupils made relatively far less errors than the epileptics in both the immediate and the deferred recognition. Both the normals and the epi- leptics made considerably more errors in the deferred recognition than in the immediate recognition. 6. The normal girls made less errors than the normal boys in both tests, but particularly in the deferred recognition, while the epileptic boys appar- ently made less errors than the epileptic girls. 7. There was no correlation between the number of mistaken recogni- tions and the chronological age of the normal children. 8. There are considerable individual differences in the ability correctly to recognize the cards and in the number of errors made. Among the epi- leptics the poorest individual made zero while the best one had nine cards correct in the immediate test, while in the deferred test the variation was from 0 to 8 cards. The errors varied from zero to five cards in the immedi- ate test, and from zero to six in the deferred. Among the normal children the scores varied from zero to ten cards in the immediate test and from zero to eight cards in the deferred, while the errors varied from zero to five in the immediate test and from zero to nine in the deferred. The amount of varia- tion in each age in the number of cards correctly recalled (the difference between the best and the poorest record) in the immediate test is as follows for the normals: Age. 7 8 9 10 11 12 13 14 15 16 17 Difference 6 9 8 3 7 3 5 4 2 3 2 9. The vast majority of the pupils interrogated based their recognitions in the deferred test merely on the memory of the pictures on the cards. Not very many pupils were certain that all their recognitions were correct. On the whole, there was a fair degree of correspondence between the subjective feeling of certainty and uncertainty and the objective results. 10. Judging by the improvement with increasing age this test would have fair value as one of a number of tests for measuring intelligence. We cannot, however, infer without further proof that one who proves to be poor in recognizing these post cards would also be poor in recognizing persons and things, although this might be so. One of the subjects in this test made a perfect score who is notorious for her ability to recognize faces. XIII. SPEED OF MOTOR REACTION Writing Circles: L5. 1. DESCRIPTION OF TEST The tapping test has usually been used as a test of motor speed. In individual experiments the taps have usually been made on a telegraph key or by means of a metal stylus on an electrically connected metal plate. In group experiments dots have been made by means of a pencil on plain sheets of paper or on paper divided into 100 1-cm. squares, in each of which the subject has been instructed to place one dot. Sometimes vertical strokes have been made instead of the dots. We decided not to use the dot method, because in a brief preliminary test we found that some epileptics beat so hard as to break the graphite, while sometimes the dots were superposed so that it was difficult to count them (we preferred not to use the cross-section paper, as this would introduce a complicating factor). Moreover, it seemed to us desirable to use a more complicated manual movement than that in- volved in making a straight line, and so we had the subjects make circles. Making circles or curvilinear lines requires a higher degree of motor coordin- ation than merely drawing straight lines. It gives a better combined measure of motor speed and coordination, we believe, than making strokes. However, we took pains to advise the subjects that no effort should be taken to make beautifully curved or perfect circles, as may be seen in the instruc- tions. The methods of making strokes and making circles with a lead pencil are both subject to the difficulty that the extent of the manual movement cannot be regulated, as is the case when the telegraph key is used. Given two subjects of equal motor ability (at least qua speed), the one who makes the larger movements will necessarily make the smaller number of move- ments. There was considerable variation in the size of the circles made by some of our subjects. Most of those who made large circles were among our young or low grade subjects. Their scores were usually low. It is, of course, possible that their scores would have been just as low or lower had they made smaller circles, as the smaller movements might have involved more difficult motor coordinations and thus have required more time for their execution. 2. DIRECTIONS FOR GIVING THE TEST "Just as soon as you turn your papers over. I want you to start writing O's, or making little circles like these (illustrate by making small circles on the board). Don"t make the circles too large; make them about the size that you make an ordinary O. Don't make half circles; try to close them. 157 158 The Measurement of Mental Traits in But don't spend any time trying to make the circles look pretty or perfectly round or go over them a second time to make them better, because you must write just as many as you possibly can. Begin at the'upper left corner, and write toward the right (indicate direction of movement on sheet of paper). Don't run the O's together. Keep them separate. You will get only a short time, so hurry and make just as many circles as you can." "Now, ready!" "Now, stop! Turn your papers. Sign your names." (Time allowed: 60 seconds.) Scoring: Record the number of circles made. No credit was given for curved lines which were less than half circles, but not many such lines were made. Circles were sometimes run together, but there was no difficulty in accurately counting interlaced circles. Normal and Epileptic School Children 159 3. TABLES. TABLE 36. Manual Movement Test (Writing Circles): L5. Normal Children No. Age Sex May 1NO. i 7 1 r>Oy Do . 0 o I 7 C irl 79 R 4 7 1 Av6 . 70 R IV . D 7 Q O 7A_ 9 /ft . Z 6 Q O fiQ fi 10 Q O Ave . Q9 O OZ . U o o Q y r>oy 7° fi :o 0 A 4 Q y 74 V 7 7 Q y Ave . 1A C\ IH . U 4 i n IU r>oy Qfi n yo . u a in P irl so r oil . O 7 i o 1U Ave . ftQ ft oy . O o O 1 1 r>oy Bo . D Q O 1 1 1 1 Pirl qr n I I 1 1 1 1 Ave . yo l b 1 9 we. Boy 100 . 0 1 9 119 fi Q o 1 9 Ave . 100 . o 4 13 Boy 103. 2 13 Girl 118. 6 13 Ave . 1 Oft 4 14 ooy 191 7 14 Girl 124. 11 14 Ave 123. 2 15 Boy 133. 2 15 Girl 109. 4 15 Ave. 121 2 1 16 Boy 105. 3 16 Girl 113.3 4 16 Ave. 111.2 1 17 Boy 103 2 17 Girl 124. 3 17 Ave. 117. 94.4 100.6 Average Boys and Girls — 97.1 Epileptics OU. Age Sex May No. ol 10 rsoy 75. 1 09 1 0 70. 1 CiA 1U4 1 0 1U ijirl 87. Ave . 78.5 QQ yy 1 1 11 99. 7Q IV 1 9 r>oy 71. QA i>4 1 9 14, Boy 113. Q1 yi 1 9 Boy 83 Ave . 89. 1 Oft 1 9 28. yo 1 "5 10 £50 y 115. B£ OO 1 Q lo Boy 38. Ave ' 76.5 07 y / 1 0 lo (jirl 87. 1 A 14 Boy 85. ey l /i 14 Boy 95. Ave . 90. so OU 1 R ID r>oy 95. 09 1 R 10 coy 68. ftQ oO 10 r>oy 61. Ave . 74.6 1 OQ lUO 15 (jirl 71. 70 /O lo Boy 92. 88 16 Boy 107. 85. 16 Boy 95. Ave . 98. 96 16 C\y\ \j\X\ 22. 98 16 ' irl 94. 106 16 Girl 74. 100. 16 Girl 00 . Ave 63.2 "87 17 Boy 84. 90. 17 Boy 116. Ave. 100. "95 if Girl 26. 105 18 Girl 92. 107 21 Girl 91. 101 24 Girl 33. Average 87. Average Girls . . . 66.9 Average Boys and Girls 77.6 4. RESULTS (See Table 36) The normal pupils averaged 97.1 circles while the epileptics averaged 77.6. The normal pupils made on the average 1.6 circles a second, and the epileptics 1.3. The epileptics did 80% as well as the normal pupils. The epileptics 160 The Measurement of Mental Traits in on the average did a little better than the nine-year group of normals, and made about 39 circles less than the oldest group of normals. The normal boys did 93% as well as the normal girls, making on the average 6.2 circles less per minute than the girls, while the epileptic girls did only 77% as well as the epileptic boys, making on the average 17.1 circles less a minute than the boys. Among the normal children the girls did better in eight ages while the boys excelled in three ages, but among the epileptics the boys excelled in four ages and the girls in two ages, no comparison being possible in the other ages. The rapidity of making circles among the norrfcal pupils increased in eight ascending ages and decreased in two ages, namely, in ages 15 and 16. In age 15, 75% of the children were classed as bright, while in age 16, 75% were classed as dull and average. In this test the highest average was made by the 14-year-old children. The gain made by the oldest group over the youngest group amounted to 46.5 circles, or an improvement of 40%. 5. CONCLUSIONS 1. In speed of motor reaction, as measured by the writing of circles, the epileptics did 80% . as well as our normal subjects. The epileptics on the average did somewhat better than the nine-year-old normal pupils. Our results are not in harmony with the conclusions of Smith, who says that the speed of tapping is not different in epileptic and normal individuals. 2. The normal girls reacted slightly more rapidly than the normal boys, while the epileptic boys reacted noticeably more rapidly than the epi- leptic girls. Earlier experiments with the tapping test have given discrepant results. With the Seguin form-board, which is a more complicated motor test than writing circles, or tapping, the boys have proved to be more rapid than the girls, especially during the first trial.1 3. The speed of motor reaction, in the form in which it is here meas- ured, increases with ascending chronological age. at least to the age of 14. It is possible that the increase would have extended beyond 14 had more sub- jects been tested. In the tapping test the increase has been found to extend to age 18 by several investigators. 4. There are considerable individual differences even in this simple test. The difference between the lowest and the highest individual scores among the epileptics amounted to 94 circles (from 22 to 116), and among the normals to 102 circles (from 40 to 142). The difference between the lowest and the highest individual scores (number of circles) in each chrono- logical age for the normal subjects was as follows: Age.. 7 8 9 10 11 12 . 13 14 15 16 17 Difference 20 43 62 59 53 31 40 42 45 . 32 35 !J. E. Wallace Wallin. Psycho-Motor Norms for Practical Diagnosis, 1916, 54ff., 69, 71. Normal and Epileptic School Children 161 These significant differences in so simple a group test are explicable on the assumption that the rapidity of making circles varies with two factors, the one general and the other specific. It varies (a) more or less with the gen- eral mental capacity of the individual and (b) with the facility with which the individual is able to execute the necessary manual movements, which, in turn, varies with innate motor capacity and practice in rapid writing. 5. In spite of its simplicity this appears to be a useful test for measur- ing differences in motor capacity, and mental ability as well, as indicated by the increasing ability shown in the test with increasing chronological age among our normal children. M T 11 XIV. GENERAL COMPARISONS AND CONCLUSIONS We present graphically on pages 162 and 163 the comparative efficiency and the comparative improvement (based only on the absolute units), of the normal and epileptic subjects, based on the group averages for all the sittings. Psychological Tests of Normal and of Epileptic School Children Immediate Memory (Digits J: AV.EF AVMO.IM Remote Memory (Digits) AV.EF a £ 416% tO £ | 25.9% Ui 5 I 1 S MS8 o N E id N E ui N io.2 - ui ■ ^ o N E ui Spontaneous Association: AV.EF AVMO.IM Addition (Correct only.). AV.EF. AVMO.IM !> rri 45.1% % * 1 35.% gj jj N E u| N E ui N 14.7% , W | «•/. £ -.2% .35% N E ui N E Addition CCorrect and Incorrect ) . AV.EF. AVMO.IM. Ardonym Test AV.EF. AVMO.IM. N E ui N E ui * 0 39.1% ^ g N E ui N E ui Tkrception : AVER AVMO.IM. Observation - Objects (immediate) : AV.EF. AVMO.IM ■ ^/k s 1 1 8 2 -5* 0 N E ui N E w 45.8* jg 1 3 °i% 1 i N E ui N E ui N— Normal. £-Epi!eptic. Av. Ef— Average Efficiency. Av. Mo. Im.— Average Monthly Im provement. E. C. Ef .- Epileptics' Comparative Efficiency. E. C. Im.— Epileptics' Comparative Improvement. 162 Normal and Epileptic School Children 163 Psychological Tests of Normal and of Epileptic School Children Observation. - Objects (Remote ) : AVEF Paired A.ssocia tes — Immediate Memory: AVEF. AVMO.IM. 23.1% ft ■ 9.7% w. 1 ■ u N . E w' 68 1% H £ I f) « 1 25.S%W g 1 1 0 2* 0 NEW NEW Paired dissociates - Remote Memory AVEF X 22.1% jjj 1 7s% s NEW Imagination ; Blots : AVEF. in 47% 2:4% N E w PV&rvf Construction AVEF ..13.0% 13.6* 12.% 5.6% 13.4% 10.3% 4.9% 8.% 2.8 blots 3.8 words (II). 46.3 40.2 54.2 67.8 72.7 28.5 44.2 76.5 12.2 10.7 The importance of the question of individual differences in psychological experiments was emphasized in one of our earliest experiments, carried out in 1901. 3 and we have been forced to recognize the question in nearly all our subsequent investigations.4 During the last decade or more many of the ablest psychologists and educationists have devoted a large amount of their productive energy to the investigation of the extent, causes and consequences of individual differences. Three practical consequences follow from the demonstrated facts regard- ing individual differences. First, we cannot hope to establish reliable age standards in mental tests by the examination of a few subjects in each age. We need to test many subjects; the more we test the smaller will the mean variation become. Second, it is not safe to attempt to guage the mentality of a subject by giving only a few tests, unless the abnormality is so pro- nounced that we cannot fail to recognize it by means of a few tests. Ideally we should attempt to explore all the fundamental mental functions or traits, but these functions have not yet been charted. The number of tests used in this investigation are none too numerous. We would increase them rather than decrease them. Third, we should determine the amount of variation in any trait from the established age (or stage) norm which may be considered abnormal or pathological, and the amount of variation required in various traits to constitute different degrees of mental deficiency (e. g., feeble-mind - edness. borderlinity, backwardness, brightness, precocity, etc.). 9. The utility of this series of tests (with individual exceptions) for the purpose for which they were designed, for use with children with mentalities of seven or eight and over, has been shown. ■Combined scores. -Accuracy scores. ^Optical Illusions of Reversible Perspective, 1905, pp. 111. 115. 120, 123, 136, 178, 227f. 309. ••Spelling Efficiency, 1911; see "variation" under "spelling efficiency;" Dental Cosmos, 1912, p^^f and p. 454f Experimental Studies of Mental Defectives, 1912, 37f, 42, 59f. Psycho-Motor Norms for Practical Diagnosis, 1916, 89f. Test av. J . 1.6 sentences . 4.6 sentences K 2 cards correct 9 cards correct L 20 circles 62 circles 174 The Measurement of Mental Traits in The value of the tests for measuring the relative mental efficiency of groups of children has been shown, first, by the fact that the efficiency tends to increase with increasing maturity (or chronological age) among the normal children. This increase is marked in view of the limited number of subjects and the unequal distribution of backward children in the different ages. From the point of view of the consistency and the size of the age increase all the tests seem satisfactory except the ink-blot test, and possibly the memoriz- ing of digits and the recognition of post cards. In the second place, while all the tests were so difficult, as given, that the oldest or ablest normal chil- dren were unable to make a perfect score, yet practically all the tests were also so easy that most of the youngest normal children were able to score. A few of the tests, however, were too difficult for a considerable number of the epileptics and a few of the normals — e. g., the memorizing of digits, paired associates, antonyms and sentence construction. Some tests are sub- ject to incidental weaknesses, e. g., the scoring in memorizing digits and apprehending objects. The value of the tests for measuring mental growth and improvement rests upon the assumption that all the tests in the same set are equal in diffi- culty. We have given some facts in the earlier pages in support of the cor- rectness of this assumption, so far as concerns a considerable number of the tests, while we have pointed out the incompleteness of some tests (e. g., word and sentence construction), the obvious lack of uniformity in one test (ink-blots) and the difficulty of making uniform series in another test (antonyms). If we may judge by the distribution of the best and the poorest scores in the different sittings, the tests seem measurably uniform, for if they are equal we would naturally suppose that the preponderance of the poorest scores would come in the early sittings and most of the best scores in the later sittings. As a matter of fact, confining the analysis to tests A to G, the poorest score by the normal pupils was made in the first sitting in five tests, in the second sitting in two tests and in the third sitting in only one test ; while the epileptics made the lowest score in the first sitting in four tests and in the second sitting in two tests. On the other hand, the best score by the normals was made in the third sitting in only one test, in the fourth sitting in four tests and in the fifth sitting in two tests; and by the epileptics in the first and third sittings in only one test each, in the fourth sitting in two tests, and in the fifth sitting in three tests. We believe these figures show that the arrangement is fairly satisfactory. As we have already explained, however, the tests can be legitimately used to measure the com- parative improvement of different groups of individuals, even though the series are not absolutely uniform in difficulty. In the construction of this system of tests the distinctive purpose was to provide a half dczen variant forms for each test, in erxh of which the Normal and Epileptic School Children 175 materials, or constituent parts were to be differently arranged, but each of which, nevertheless, was to be of approximately the same degree of difficulty. In other words, the aim was to construct a set of uniform or unvarying scales for measuring the degree of mental growth or change which different sub- jects, such as the feeble-minded, epileptic, psychotic, backward, normal, or supernormal may undergo during various intervals of time, either as a result of the natural growth impulse, or as a result of various forms of treatment, training or handicap. The tests previously employed for this purpose had made u:e of identical test materials in each successive set, in consequence of which the results have been unduly affected by the factors of memory and familiarity, especially when the tests have followed each other at close inter- vals. Between the time of the devising of these tests and the publication of this study many group intelligence tests have been constructed in two or more variant forms of putative equal difficulty in which the components have been differently arranged. Before this manuscript was completed a brief summary of the investiga- tion and of the theories of epilepsy was published in the Problems of Sub- normality, 1917, pp. 350-381. I have been aided in the proof reading by Miss Mildred Rothhaar and Miss Charlotte Fiala. I am under obligations to the World Book Company for the use of the graphs on pages 163 and 164. PRISS OHIO BTA1 ft Mm